on the date of the nyâyâvatâra - piotr balcerowicz · 2016-11-07 · on the date of the...

Published in: On the Understanding of Other Cultures. Proceedings of the International Conference on Sanskrit and Related Studies to Commemorate the Centenary of the Birth of StanislawSchayer (1899–1941). Warsaw University, Poland, October 7–10, 1999. Ed. by Piotr Balcerowicz and Marek Mejor. Studia Indologiczne 7 (2000) [Warsaw, Poland]Reprinted in: Essays in Indian Philosophy, Religion and Culture, Ed. by Piotr Balcerowicz and Marek Mejor. Motilal Banarsidass, Delhi 2004.

Essays in Indian Philosophy, Religion and Literature, pp. 331–370.

On the Date of the Nyâyâvatâra

PIOTR BALCEROWICZ

The Nyâyâvatâra (NA.), a work in thirty-two verses—and hence also calledDvâtriôœikâ—erroneously ascribed by tradition to Siddhasena Divâkara, is deemedto open a new era in the history of Jaina epistemology. It is mostly in the realm ofJaina epistemic pursuits that the Dvâtriôœikâ might claim the status of an innovativeor prototypical work. When we, however, consider the development of logic andepistemology in India on a larger scale, the work seems to lose its flavour oforiginality and novelty.

The problem of the exact dating of the Nyâyâvatâra1 should be solvedindependently, irrespective of whether Siddhasena Mahâmati, the author of theNyâyâvatâra, can be accurately identified with a Siddhasena (Divâkara?), thecelebrated author of a series of Dvâtriôœikâs, or— inaccurately—with SiddhasenaDivâkara, the author of the Saômati-tarka-prakaraòa.2 The earliest indication of the

1 The most comprehensive bibliographic survey of publications on Siddhasena

Divâkara is furnished by UPADHYE (1971) in his ‘Introduction’ (pp. xi-xxvii) and‘Bibliographic Survey’ (pp. *3–*72). A few more publications have been publishedsince the review: MOOKERJEE (1971), DHAKY (1981–82), GRANOFF (1989–1990),DHAKY (1990), DHAKY (1995) and WAYMAN (1996).

2 The question has been discussed at length in BALCEROWICZ (2001) and BALCEROWICZ

(forthcoming). A number of features found in the Nyâyâvatâra and the Saômati-tarka-prakaraòa evince a discrepant attitude towards the Jaina Canon and tradition of both works,different choice of vocabulary, which was not necessitated by the use of different languages(Sanskrit and Prakrit), and the notions and ideas they use are likewise at variance. A detailedcomparison of the contents, style and philosophical background of Siddhasena Mahâmati’sNyâyâvatâra and Siddhasena Divâkara’s Saômati-tarka-prakaraòa (vide infra p. 361 f.),brings me to the conclusion that these two works were written by two different persons.Following the findings, esp. in view of the lack of any hint that the author of STP. knew ofDiónâga, I would maintain that STP. must have been composed slightly before or circa 500C.E. Comp. also MUKTHAR (1948): ‘The following points are clear: (1) The Dvâ.s were notcomposed in the present order; (2) they are not of one and the same Siddhasena; (3) theNyâyâvatâra is one of them; (4)… No indisputable evidence is brought forth for thecommon authorship of the Dvâ.s, Nyâyâvatâra and Sanmati. … There were thus three clear

332 PIOTR BALCEROWICZ name of the author of NA. is, it seems, Haribhadrasûri who makes mention of himunder the appellation ‘Mahâmati’ (vide infra p. 361). The subsequent source fromwhich we learn that NA. was composed by a Siddhasena is NAVV. of Œântisûri. Theauthor of NA. is explicitly identified there in at least four places3. In the lastreference Œântisûri is even more specific to give the full name of the author as well:Siddhasenârka4. All other references we encounter in the Jaina literature of thisperiod are to Siddhasena Divâkara as the author of other works, but none to him asthe author of NA.5 Strangely enough, the available colophons of NA. and NAV.

and distinct Siddhasenas: (1) the author of Sanmati; (2) the author of Nyâyâvatâra; (3) andanother, the author of some Dvâ.s.’ [according to UPADHYE (1971:*28)].

3 [1] NAVV.36§ 7 (p. 95.8): œrîsiddhasena-ghaþita-sphuþa-gîÿœalâkâô œuddhâmavâpya vimalaô vihitaô mayÎtat // , [2] NAVV.21.§ 2 (p. 78.9–10): evaô-rûpasyavadhaÿ tyâgaÿ siddhasenârkasyêty arthaÿ., [3] NAVV.53.§ 2 (p. 107.18):siddhasenasya sûtra-kartuÿ…, [4] NAVV.1.§ 11 (p. 13.14–15): tat kiô svâtantryeòa?na ity âha—siddhasenârka-sûtritam iti. siddhasena eva jagaj-jantu-mano-moha-saôtatitâm asîtamaÿ samûhâpoha-kâritvât arka iva arkaÿ, tena sûtritam.

4 Indeed, Œântisûri speaks of Siddhasenârka, not of Siddhasena Divâkara, asUPADHYE (1971: xxiv) would like to have it: ‘iii) … The earliest author, as far as Iknow, who specifies the name of Siddhasena Divâkara as the author of the Nyâyâvatârais Œântisûri of the 11th century A.D. or so.’ Nevertheless, I would side with UPADHYE inasserting that ‘Arka’ is just another name for ‘Divâkara’. Thus, I see absolutely nojustification for the contention of DHAKY (1995: 49, n. 9), who—commenting on theclause: siddhasena eva jagaj-jantu-mano-moha-saôtati-tâmasîtamaÿ-samûhâpoha-kâritvât arka iva arkaÿ, tena sûtritam—maintains that ‘“Arka” here is not in the sense of“Divâkara” but “essence”.’ Clearly, arka here is an epithet of Siddhasena, who iscompared to the sun (arka iva; arka=divâ-kara), and by him (tena) the idea discussedbefore is composed in a sûtra form (sûtritam). If we took arka as DHAKY would like it,the whole clause could hardly be construable.

5 Cf., e.g. [1] UPADHYE (1971: xiii): ‘Haribhadra is one of the earlier authors tomention Siddhasena Divâkara and his Sanmati. First, he calls him Œrutakevalin; andsecondly, he tells us that his name was Divâkara (p. *1). Then he has a pun on the namethat he was like Divâkara, sun, to the darkness of Duÿšama-kâla.’ [2] UPADHYE (1971:xvii): ‘…Pûjyapâda …quotes Siddhasena’s Stuti III.16 in his Sarvârthasiddhi (II.10;VII.13).’ [3] H. R. Kâpadîâ (AJP., ‘Introduction’, Vol. II, pp. 98 ff.): ‘Haribhadra refersto Siddhasena in his Sammaipayaraòa, in his Anekântajayapatâkâ as well as in hisPaôcavatthuya (vv.1047–8), calling him Suyakevali. Jinadâsagaòi (c. 676 A.D.) refersto him thrice in his Niœîhavisehacuòòî.’ [4] DAVE (1962): ‘So Siddhasena is earlier thanMallavâdi and the tradition puts him as a contemporary of Vikramâditya who flourishedin 57 B.C.’ [according to UPADHYE (1971:*53)] and ‘Akalaóka and Vidyânanda quotethe Sanmati.’ [according to UPADHYE (1971:*59)].

ON THE DATE OF THE NYÂYÂVATÂRA 333 contain no reference to the name of the author of NA. Therefore, for the sake ofconvenience, I tentatively call the author of the Nyâyâvatâra Siddhasena Mahâmati.6

On subsequent pages I shall try to establish the chronology of the Nyâyâvatâraand its correlation to other, mostly Buddhist, works.

There is a variance of opinions regarding the date of Siddhasena as the author ofthe Nyâyâvatâra and the date of the Nyâyâvatâra, and these fall in four groups:(1) Siddhasena was pre-Diónâgan7, (2) Siddhasena flourished soon after Diónâgaand before Dharmakîrti8, (3) Siddhasena belongs to a post-Dharmakîrtian tradition9,whereas (4) some hold that his date is still an open question10.

The first to notice some chronological dependence of NA was JACOBI (1926: iii),who observed that ‘To about the same time [i.e. Œaka-year 598 = 677 C.E.—P.B.]belongs Siddhasenadivâkara whom Haribhadra quotes; for he uses, no doubtDharmakîrti1, though he does not name him.’ There are two points, according tohim, that justify such a conclusion: (1) Siddhasena (NA.5) applies the termabhrânta—and Dharmakîrti was the first to use it in his definition of pratyakša, thusimproving upon Diónâga’s definition11—to both pratyakša and anumâna;(2) Siddhasena (NA.1112) ‘extends the distinction of svârtha and parârtha, whichproperly applies to anumâna only, to pratyakša also, ibidem 12 f. Apparently, hethought to improve on Dharmakîrti by a wholesome generalisation of nicedistinctions!’ (JACOBI (1926: iii, n. 1)). VAIDYA (1928: xviii-xx) elaborates uponJACOBI’s laconic remarks. (3) He further brings up one more important point:

‘…verses 6 and 7 above of Nyâyâvatâra unmistakably presupposeDharmakîrti and the later phase of the Yogâcâra school, as, withoutthem, it is difficult for us to explain why Siddhasena Divâkara isrequired to these views and emphatically declare:

6 See BALCEROWICZ (2001) and BALCEROWICZ (forthcoming).7 E.g. SUKHLAL (1945/a) and SUKHLAL (1945/b), H. R. Kâpadîâ (AJP.,

‘Introduction’, Vol. II, pp. 98 ff.), KRAUSE (1948), DAVE (1962), SUKHLAL–DOSHI

(1928), WILLIAMS (1963: 19), MATILAL (1985: 241).8 E.g. Malvania (NASV., ‘Introduction’, pp. 141 f.) and QVARNSTRÖM (1999: 178).9 E.g. JACOBI (1926), VAIDYA (1928), MUKTHAR (1948), V.P. Johrapurkar

(‘Introduction’ to VTP., pp. 41 ff.).10 UPADHYE (1971: xxv).11 PS.1.C,k3c-d: pratyakšaô kalpanâpoðhaô nâma-jâty-âdy-asaôyuktam, and

NB.1.4: tatra pratyakšaô kalpanâpoðham abhrântam, respectively.12 Not NA.12, as VAIDYA (1928: xviii, line 16) has it.

334 PIOTR BALCEROWICZ

sâkÂâlâpRâitâwasâSyâ wRaNtâTvaisâiÏâtâ: SföÂþâmµ|pRâmaÒaâÔ SvaNyâinâ{aaiyâ ÝâyâisâÏO pRâisâZyâitâª 7ª ’ (p. xx).

On my part, I would only add that also NA.31 (pramâtâ svânya-nirbhâsî) must havebeen inspired by similar thoughts as NA.7.

The first argument is sound. Admittedly, the idea of abhrântatva of perception waslatent in pre-Dharmakîrti’s literature13, but Dharmakîrti was the first to use the term.14

13 The term bhrânta (bhrânti) itself is attested in the pre-Dharmakîrtian literature, for

instance in MAVBh.1.4; SacAcBh.(2).2 and in several places of MSA. and Comm.thereon (11.13a, 11.15, 11.17, 11.24–26, Comm. ad 10.2 and ad 11.27). In all suchpassages, however, the term bhrânti does not occur in the context of pramâòa, still lessof valid perception (pratyakša), in the first place. The term refers either to a generalerror based on the perceiving of subject–object duality in the world (dvaya-bhrânti), andis synonymous to mâyâ (in MSA.), or to the nature of the cognised object(MAVBh.1.4). The latter rests on a rather subtle difference: erroneous is not thecognition as such (the emphasis on the inner, cognitive aspect, viz. erroneouscorrespondence of an act of cognition), but the way an object is constituted in thecognition (the emphasis on the ‘outer’, ‘objective’ side). In none of these texts where wecome across the term bhrânti is the idea of cognitively valid procedures (pramâòa)discussed; at the most, it points to an antithesis of a general, soteriologically relevantoutlook of a person, viz. citta-bhrânti / bhrântaô cittaô / kšipta-cittaô (SacAcBh.(2).2).This is confirmed by the application of the past passive participle to people MSA.11.18(loko hy abhrântaÿ), as loci of bhrânta-citta. To sum up, none of the occurrences of theterm (a)bhrânta in Yogâcâra works seems to have been an inspiration for Siddhasena.

14 An interpretation of non-erroneousness (abhrântatva) is offered by Dharmottara inNBÞ.3.2: abhrântam artha-kriyâ-kšame vastu-rûpe ’viparyastam ucyate. artha-kriyâ-kšamaô ca vastu-rûpaô sanniveœôpâdhi-varòâtmakam. tatra yan na bhrâmyati tadabhrântam. Thus, there are two pivotal aspects of abhrântatva, the lack of contrariety(aviparyastatva) and its reference to a thing capable of efficient action (artha-kriyâ-kšama-vastu). The first element, viz. the lack of contrariety, or correspondence to facts,may be taken to have been preconceived in the idea of avyabhicâritva in the non-Buddhist literature as early as NS.1.1.4 (indriyârtha-sannikaršôtpannaô jñânamavyapadeœyam avyabhicâri vyavasâyâtmakaô pratyakšam) and NBh. ad loc. (yadatasmin tad iti tad vyabhicâri, yat tu tasmin tad iti tad avyabhicâri pratyakšam iti). Thistendency can be also observed in Jaina sources, e.g. in TBh.1.32 (p. 30.6, p. 31.1–2):jñâna-viparyayo ’jñânam iti … mithyâ-darœana-parigrahâd viparîta-grâhakatvametešâô (= viparyayânâm). tasmâd ajñânâni bhavanti. See also PVin.I(1).4 (p. 40, n. 1).

However, the second element artha-kriyâ-samartha, the capability to execute efficientaction, is Dharmakîrti’s innovation, see PV.1.3: pramâòam avisaôvâdi jñânam artha-kriyâ-sthitiÿ / avisaôvâdanam œâbde ’py abhiprâya-nivedanât // , as well as PV.2.3:

ON THE DATE OF THE NYÂYÂVATÂRA 335 If we were to take this latency of abhrântatva as a serious counter-argument—whichseems totally unconvincing to me—it would imply that Siddhasena chose the termabhrânta intuitively and applied it indiscriminately to both pratyakša and anumâna, asif he had adumbrated that Dharmakîrti would once qualify perception as a non-erroneous cognition and inference as erroneous. Coincidentally, his way of expressionwould tally with the actual term used by Dharmakîrti, who would follow him!

Whereas I completely agree with JACOBI–VAIDYA’s first argument, their secondargument is not entirely convincing to me. Indeed, we cannot understand the idea ofsvârtha-vâkya and parârtha-vâkya (NA.10) as well as svârtha-pratyakša andparârtha-pratyakša (NA.11) without Diónâga’s and Dharmakîrti’s well-knowndivision of svârthânumâna and parârthânumâna. It should suffice to remind thereader of NB.2.1–2: /1/ anumânaô dvividhâ. /2/ svârthaô parârthaô ca. and ofNB.3.1–2: /1/ tri-rûpa-liógâkhyânaô parârthânumânam. /2/ kâraòe kâryôpacârât.

There can be no doubt, in my opinion, that the idea of vâkyasya parârthatvaônecessarily presuposses the idea of vâkyasya svârthatvaô, and that both theseconcepts—expressed in NA.10 ff.: parârthaô mânam âkhyâtaô vâkyaô—couldonly have been developed in the context of svârtha-° and parârthânumâna.

However, the idea of svârtha-pratyakša and svârthânumâna / parârthânumâna—central for JACOBI–VAIDYA’s reasoning—which is a prerequisite for Siddhasena toestablish the thesis of parârtha-pratyakša, predates Dharmakîrti and is found also inthe saógraha-œloka (iti œâstrârtha-saógrahaÿ) of Nyâya-praveœa of Diónâga’sdisciple, Œaókarasvâmin15.

artha-kriyâ-samarthaô yat tad atra paramârtha-sat / anyat saôvåti-sat proktaô te sva-sâmânya-lakšaòe // . Cf. also HATTORI (1968: 14): ‘The concept of “artha-kriyâ” isunfamiliar to Dignâga, but it is an important criterion for the distinguishing of “sva-lakšaòa” from “sâmânya-lakšaòa” in Dharmakîrti’s system of thought.’, HATTORI

(1968: 79 § 1.14) and FRANCO (1987: 445 n. 203), esp.: ‘the concept of arthakriyâ doesnot appear anywhere in Dignâga’s writings.’

On the other hand, Dharmakîrti’s definition does go back to the usage of the termbhrânta in the Yogâcâra school; his novelty was to mould it to the demands of hispramâòa theory. We should remember about his idea that anumâna is bhrânta (cf.PVin.II(1).2.6–7, p. 24.6–7: de ma yin la der ÿdzin phyir || ÿkhrul kyaó ÿbrel phyirtshad ma ñid || = atasmiôs tad-graho bhrântir api sambandhataÿ pramâ // ), which is aproper point of reference here, not the nature of citta / vijñâna.

15 sâdhanaô dûšaòaô cÎva sâbhâsaô para-saôvide / pratyakšam anumânaô casâbhâsaô tv âtma-saôvide // . As for a possible objection that the verse might be later,at least the commentator Haribhadra takes it to be a genuine part of the original work(NP.(1). p. 9.12 ff.): …ity âdâv eva œlokaÿ, etc.


Certainly it is not Œaókarasvâmin who developed the concept of svârtha-° andparârthânumâna. It was probably fathomed by Vasubandhu16, but terminologicallyconceived and worked out by Diónâga17.

There are, however, two other traits to be noticed in the aphorisms NA.10–11 thatpoint, in my opinion, to Dharmakîrti as their source. The first of them is the idea ofmetaphorical transference (upacâra) used in a very similar context. Siddhasenaspeaks of the term ‘inference for others’ (parârthaô mânam)—which denotes, to beprecise, the inferential cognition arisen in another person as a result of anargumentative procedure and / or debate—which is applied through metaphoricaltransference (upacâra) to a ‘syllogistic’ sentence (vâkyaô), that is, as a matter offact, merely a cause of such a cognition in another person. We come across the sameidea in NB.3.1–2, and even the wording is to a certain extent similar (one shouldhere take into account stylistic differences necessitated by the succint sûtra styleand versified kârikâs). Below, both passages in question are given for the sake ofconvenience, relevant expressions being underlined:18

/NA.10/ sva-niœcayavad anyešâô niœcayôtpâdanaô budhaiÿ /parârthaô mânam âkhyâtaô vâkyaô tad -upacârataÿ //

NB.3.1–2: tri-rûpa-liógâkhyânaô parârthânumânam. kâraòe kâry ôpacârât.

Dharmottara’s gloss confirms the interpretation: kâraòa = vacana = vâkya; kârya= [parârtha] anumâna = parârtha-mâna.19 Siddharšigaòi, whose comments are in asimilar spirit, even quotes a verse, that establishes the relation between speech(kâraòa) and resulting cognition (kârya)20. Another striking feature is that the

16 Cf. FRAUWALLNER (1933: 476–477 [297–298]) and HATTORI (1968: 12, n. 60).17 E.g.: PS.2.1ab: anumânaô dvividhâ svârthaô tri-rûpâl liógato parârthânumânaô

tu sva-dåšþârtha-prakâœanam; also PS.3.1ab, and PSV.2, K 109a.2–3 = V 27a.5(svârthânumâna): tshul gsum paÿi rtags las rjes su dpag paÿi don (V: rjes su dpag parbya baÿi don) mthoó ba gaó yin pa de ni raó gi don gyi rjes su dpag paÿo. Cf. RANDLE

(1926: 28–9), HATTORI (1968: 78, n. I.11) and Steinkellner’s note 1, p. 21 in PVin.II(2).18 In my analysis throughout I shall use, for the sake of brevity, underlining to mark

corresponding phrases in NA. and works of Dharmakîrti or of other authors.19 NBÞ.3.2, p. 150.12–151.1: tasmin kâraòe vacane kâryasyânumânasyôpacâraÿ

samâropaÿ kriyate. tataÿ samâropât kâraòaô vacanam anumâna-œabdenôcyate.aupacârikaô vacanam anumânaô, na mukhyam ity-arthaÿ.

20 vikalpa-yonayaÿ œabdâ vikalpâÿ œabda-yonayaÿ /kârya-kâraòatâ tešâô nârthaô œabdâÿ spåœanty api //

The verse is so far untraced, but it refers most probably to another verse byDharmakîrti, and—at any rate—to an idea expressed in PV.1.286.

ON THE DATE OF THE NYÂYÂVATÂRA 337 reference to the idea of the metaphorical transference (upacâra) occurs in bothworks (NA. and NB.) precisely at the moment of introducing the discussion ofparârtha-anumâna and that this is the only occurrence of this idea in both works.Neither the term nor the idea as such is encountered at any other point.

Further, VAIDYA’s third point is rather weak, as well. He says: ‘These verses[NA.6–7—P.B.] contain the favourite view of the Yogâcâra School on the subjectand the object (grâhya and grâhaka), which both they declare illusory. This view isdependent on the definition of pratyakša, and though its origin must be sought inDiónâga’s works [emphasis—P.B.], the scholars who brought the idea to perfectionare Dharmapâla and his pupil Dharmakîrti.’ (p. xix). Indeed, we find the ideareferred to by VAIDYA in Diónâga’s works, e.g. in his PS.1.1021 (cf. n. 67):

Vasudhararakšita / Seó-rgyal 15b.4: Kanakavarman / Dad-paÿi œes-rab 96a.4–5:

gaó tshe snaó ba de gšal bya |tshad ma daó deÿi ÿbras bu ni |ÿdsin rnam rig pa de yi phyir |de gsum tha dad du ma byas ||

gaó ltar snaó ba de gšal bya |tshad ma daó deÿi ÿbras bu ni |ÿdsin rnam rig paÿo de yi phyir |de gsum tha dad du ma byas ||

However, we do not find anything in NA.722 that would allude to either Diónâga’sspecific theory of triple division of vijñâna or to Dharmakîrti’s ideas23. What we dofind instead is the realist’s position that, at least: (1) acts of cognition are real,(2) acts of cognition happen to be true and accurate, (3) acts of cognition are self-validatory, (4) acts of cognition are accurate representations of external world,(5) the external world is real.

To have Diónâga’s or Dharmakîrti’s views criticised here we would need anexplicit element of sva-saôvitti (sva-saôvedana) or phala. There are three possibleexpressions in NA.7 that might refer to sva-saôvitti (and none to refer to phala):(A) sphuþaô, (B) svânya-niœcâyi, (C) dvaya-siddhau. (Ad A) The first of the list ishighly improbable, for it never—to my knowledge—is used in Buddhist sources torefer to the idea of self-revelatory character of cognition (sva-saôvitti). It isgenerally used to describe either the veracious, direct, non-inferrential or the

21 See HATTORI (1968: 107, n. 1.67):

yad-âbhâsaô prameyaô tat pramâòa-phalate punaÿ /grâhakâkâra-saôvittî trayaô nâtaÿ påthak-kåtam //

22 sakala-pratibhâsasya bhrântatvâsiddhitaÿ sphuþaô /pramâòaô svânya-niœcâyi dvaya-siddhau prasidhyati //

23 On the triple and fourfold division of vijñâna in the Yogâcâra school see HATTORI

(1968: 107, n. 1.67).

338 PIOTR BALCEROWICZ indubitable, etc., character of acts of cognition. It is explained by Siddharšigaòiaccordingly by suniœcitatayâ. (Ad B) The second of these expressions (as a similarone in NA.31: svânya-nirbhâsî) mentions in fact only two aspects: the cognition(sva) and its counterpart—the object (anya). If sva were to mean sva-saôvitti, theprimary act of cognition would not be mentioned; if anya were taken to be sva-saôvitti, the bahyârtha, so fundamental for the realist, would not be mentioned. Theexpression is explained in NAV. as sva-para-prakâœakaô, which does not bringanything new to our analysis. However, it is coupled in the NAV. with the thirdexpression. (Ad C) The third expression refers to duality, in the first place, and—like in the preceding case—it is highly problematic to take it to allude to Diónâga’stheory of triple division of vijñâna. However, it is Siddharšigaòi’s gloss on dvaya-siddhau: svarûpârtha-lakšaòa-yugma-nišpattau that could be implicative of sva-saôvitti: in it, svarûpa might refer to the self-revelatory character of cognition. It isespecially suggestive in view of his statement in NAV.1 quoted in n. 24. But eventhen, the third aspect of an act of cognition (apart from the act as such, an object),its self-revelatory character, is not explicitely mentioned here. As a matter of fact,Siddharšigaòi, so well conversant with Buddhist ideas, would not have wasted theavailing opportunity to indicate the idea of sva-saôvitti and to utilise it24, if he hadnoticed any allusion to the Buddhist theories of triple or fourfold division of vijñânain the aphorism of Siddhasena. In other words, there is nothing in Siddharšigaòi’sgloss that might suggest that Siddharšigaòi had seen any point of convergencebetween the ideas expressed in NA.7 and certain concepts ascribable to Diónâga, asVAIDYA would like it. To expell our doubts, he concludes, as a matter of fact, withanyathâ prameyâbhâve pramâòâbhâvât, to show that no third element is implied.

In my opinion, NA.7 can be safely taken to disprove the doctrine of illusorycharacter of worldly appearance propounded by the Buddhist idealist (Vijñâna-vâdin), whose ideas directly influenced Diónâga and Dharmakîrti25. The aphorism is

24 As he does in several places, for instance in NAV.1 (the section beginning with:

ayam atrâbhiprâyaÿ: sva-saôvedanaô prati nikhila-jñânânâm eka-rûpatayâ sâkšât-karaòa-caturatvân nâsty eva bhedaÿ…) , NAV.29 (the section beginning with:tathôrarî-kåta-yogâcâra-matam api balâd anekânta-prakâœa-rajjur âvešþayaty, ekasyâpijñânasyâneka-vedya-vedakâkâratayâ prathanôpagateÿ…, and the section beginningwith: atha jñâna-vâdy advaita-prakâœam alakšitam abhyupetya tena bâhuvidhyaôdadhâno bodho bâdhyamânatvâd bhrânta ity abhidadhyât, tad ayuktam … ), NAV.31,etc.

25 Cf. HATTORI (1968: 106, n. 1.65).

ON THE DATE OF THE NYÂYÂVATÂRA 339 clearly evocative of such ideas as those expressed, e.g. in MAV.1.1,326, Triôœ.1,2927 or Viôœ.1ab28, 1629. As Siddharšigaòi expresses himself, if there is nothing tobe cognised, there can be no cognition. Therefore, to establish the thesis of theexistence of the external world is essential for the realist. Similarly, mutatismutandis, it is crucial for the Buddhist idealist to deny the existence of externalobject. Furthermore, NA.7 fits quite well into the line of critics of the so-called‘Dreaming Argument’30, that was commonly ascribed to the Buddhist and refuted,for instance, by Kumârila (MŒV.4.(Nirâlambana-vâda).23, p. 159.7–8),Uddyotakara (NV. on NBh.4.2.33), Œaókara (BSŒBh.2.2.5.29, p. 476.2–3) and bySiddharšigaòi (NAV.29, the Œûnya-vâda section).

Therefore, NA.7 is not a very useful indication to establish the date of theNyâyâvatâra. However, there is a number of other conspicuous traits, that areinstrumental in establishing the time of composition of the treatise quiteconvincingly as posterior to Dharmakîrti. To achieve this, I shall analyse severalaphorisms of NA. step by step in order to show Siddhasena’s indebtedness toDharmakîrti (especially to NB.) in respect not only of certain ‘loan’ ideas but also,partially, of the dialectical structure of the text. Some of the following points are notentirely convincing, when taken singly. Their high number, on the other hand, couldnot have been a matter of mere coincidence.

[1] The opening line of NA.0 (pramâòa-vyutpâdanârtham idam ârabhyate)closely resembles the formulations of HB. p. 1*.5–6: parokšârtha-pratipatteranumânâœrayatvât tat-vyutpâdanârthaô saókšepata idam ârabhyate.

[2] Practically, the very first ideas expressed in the opening lines of NA. and NB.are very similar and have similar wording:

26 abhûta-parikalpo ’sti dvayaô tatra na vidyate / œûnyatâ vidyate tv atra tasyâm api

sa vidyate // artha-sattvâtma-vijñâpti-pratibhâsaô prajâyate / vijñânaô nâsticâsyârthas tad-abhâvât tad apy sat //

27 âtma-dharmôpacâro hi vividho yaÿ pravartate / vijñâna-pariòâme ’sau pariòâmaÿsa trividhâ // acitto ’nupalambho ’sau jñânaô lokôttaraô ca tat / âœrayasya parâvåttirdvidhâ daušþhulya-hânitaÿ //

28 vijñâpti-mâtram evÎtad asad-arthâvabhâsanât /29 pratyakša-buddhiÿ svapnâdau yathâ sâ ca yadâ tadâ / na so ’rtho dåœyate tasya

pratyakšatvaô kathaô matam // (to be coupled with NA.6–7).30 In a typical formulation: ‘The sensation in the waking state is erroneous, because it

is a cognition, like the sensation in a dream’ (jâgrat-saôvedanaô bhrântaô,pratyayatvât, svapna-saôvedanavat). See TABER (1994).


NA.1: pramâòaô […] pratyakšaô ca parokšaô ca dvidhâ .

NB.1.1.2–3: dvividham samyag-jñânam. pratyakšam anumânaô ca.

The differences are that (1) Siddhasena does not have samyag-jñânam (thisexpression is completely absent from NA.), and (2) he has parokša instead ofanumâna. The second difference is dictated by the Jaina demand to incorporateœâbda / âgama (testimony, verbal cognition) in the pramâòas (as a quasi-separatecategory, one of the two—alongside anumâna—primary sub-divisions of parokša).However, Siddhasena marks an important, widely known shift in Jainaepistemology, to interpret the directness of pratyakša in terms of sensory organs(akša = indriya), not—as it was customary—in terms of the cognitive subject (akša= jîva / âtman). The commentators are quite explicit about the interpretational shift,which diverts from the Âgamic tradition advocated, e.g. by Akalaóka among manyothers.31

The question is what prompted Siddhasena to introduce this shift? Obviously,these were the demands of the general philosophic discourse in India to be up-to-date with and understood by such schools as Nyâya or Sâôkhya. But I am deeplyconvinced that what was responsible for that shift to take the senses as the criterionof directness of pratyakša was, to a larger degree, Dharmakîrti’s inspiration and hisworks (predominantly the Nyâya-bindu) as a groundwork for the Nyâyâvatâra. Forwhat other reason would Siddhasena speak of the two-fold division of validcognition and justify it by referring to Dharmakîrtian ideas? And thereby we cometo another element of NA.1 that resembles Dharmakîrti’s formulations.

According to Siddhasena, the factor responsible for the division of cognitive actsinto pratyakša and parokša is not the character of the ‘cognising organ’ (akša),either the sense organs (indriya) or the cognitive subject (jîva, âtman) as such. In hisopinion, what is crucial in the categorisation of pramâòas is the character of theobject of cognition that determines the way the object of cognition is determined, orcognised. Clearly, in this revolutionary rearrangement he goes against the wholeJaina tradition. The expression he uses: meya-viniœcayât (NA.1d), can hardly beexplicated without Dharmakîrti’s formulations: tasya višayaÿ svalakšaòam(NB.1.1.12), so ’numânasya višayaÿ (NB.1.1.17) and mânaô dvividhaô višaya-

31 See, e.g. NAV.1: pratyakšaô cêty-âdi; tatra siddhânta-prasiddha-pâramârthika-

pratyakšâpekšayâkša-œabdo jîva-paryâyatayâ prasiddhaÿ. iha tu vyâvahârika-pratyakša-prastâvâd akša-dhvanir indriya-vacano gåhyate. tataœ câkšaô pratigataôpratyakšaô. yad indriyam âœrityôjjihîte ’rtha-sâkšât-kâri jñânaô tat pratyakšam ityarthaÿ. … akšebhyaÿ parato vartata iti parokšam. akša-vyâpâra-nirapekšaô mano-vyâpâreòâsâkšâd-artha-paricchedakaô yaj jñânaô tat parokšam iti bhâvaÿ.

ON THE DATE OF THE NYÂYÂVATÂRA 341 dvaividhyât (PV.2.1ab)32. Even Siddharšigaòi33 confirms that what is pivotal for thedistinction is, in the first place, the character of extrinsic data apprehended by thecognitive subject, that determines the way of apprehension, irrespective of whetherthe ‘cognising organ’ are the senses or the soul. It is only in NA.4d that Siddhasenaemphasises the way of cognising, or ‘[the manner of] grasping [an object]’, for thefirst time: grahaòêkšayâ. But even then, the tradition in the person of thecommentator Siddharšigaòi takes the locution to refer to the existence and thenature of the cognoscible (see his detailed discussion in NAV.4 apropos ofgrahaòêkšayâ).

Truly, a formulation similar to NA.1 is found also in PS.(1).1.2ab34, soDharmakîrti is not the only potential source. However, if we compare what Diónâgahas further to say on the two-fold division of pramâòa, we discover that he does notmention the reason for the division explicitly, as Dharmakîrti and Siddhasena do,but he merely points to the parallelism of the two-fold prameya35. Clearly, one caneasily understand the statement lakšaòa-dvayaô prameyam to be a reason adducedby the author for pratyakšam anumânaô ca pramâòe (‘there are two cognitivecriteria: perception and inference, [because] the cognoscible has twocharacteristics’), as the commentator himself does36. However, neither the word‘because’ (yasmât / °−tvât) nor any direct explanation why pramâòa is twofold ismentioned by Diónâga explicitly, as it is by Dharmakîrti in PV.2.1bc (višaya-dvaividhyât œakty-aœaktitaÿ / artha-kriyâyâm … ). Moreover, Diónâga’s elucidation

32 Cf. also PV.2.63 (anumâna-vicâraÿ): na pratyakša-parokšâbhyâô meyasyânyasya

sambhavaÿ / tasmât prameya-dvitvena pramâòa-dvitvam išyate //33 NAV.1 ad loc. (on meya-viniœcayâd): bahir-arthaô punar apekšya kaœcic

cakšurâdi-sâmagrî-bala-labdha-sattâkaÿ svâvayava-vyâpinaô kâlântara-sañcarišòuôsthagita-kšaòa-vivartam alakšita-paramâòu-pârimâòðalyaô sannihitaô viœada-nirbhâsaô sâmânyam âkâraô sâkšât-kurvâòaÿ prakâœaÿ prathate, tatra pratyakša-vyavahâraÿ pravartate. yaÿ punar lióga-œabdâdi-dvâreòa niyatâniyata-sâmânyâkârâvalokî parisphuþatâ-rahitaÿ khalv âtmano ’rtha-grahaòa-pariòâmaÿsamullasati sa parokšatâô svî-karoti.

34 pratyakšam anumânaô ca pramâòe (Vasudhararakšita/Seó-rgyal 13b.6: móon sumdaó ni rjes su dpag tshad ma; Kanakavarman/Dad-paÿi œes-rab 94a.4: móon sum daó nirjes su dpag tshad ma dag ni); here, the difference is the absence of the explicit (sc. witha numeral, not with the dual form °−e / dag) mention of ‘twofold’ (dvividham / dvidhâ /rnam pa gñis).

35 PS.(1).1.2bc: lakšaòa-dvayam / prameyam … (Vasudhararakšita/Seó-rgyal 13b.6 =Kanakavarman/Dad-paÿi œes-rab 94a.5: mtshan ñid gñis gšal bya).

36 yasmât = Vasudhararakšita/Seó-rgyal: …phyir še na; Kanakavarman/Dad-paÿi œes-rab: … gaó gi phyir.

342 PIOTR BALCEROWICZ of the two-fold character of pramâòa (lakšaòa-dvayaô prameyam) differscompletely in character from Siddhasena’s explanation (meya-viniœcayât) andDharmakîrti’s (višaya-dvaividhyât œakty-aœaktitaÿ artha-kriyâyâm … ), insofar as itis ‘static’ or ‘taxonomic’, viz. plainly juxtaposes two cases of double division(pramâòa—prameya), whereas Siddhasena and Dharmakîrti offer ‘intentional’, i.e.actively directed to an object of cognition either by way of determining it(viniœcaya, cognitive activity) or by any sort of efficient action (artha-kriyâ).

[3] Apart from the noticeable similarity in wording between NA.1 and NB.1.1.2–3, there is another striking correspondence to be observed in the ideas expressed byDharmakîrti and Siddhasena. Siddharšigaòi introduces NA.1 in the following way:tatra tâval lakšaòa-saôkhyâ-vipratipattî nirâcikîršur âha. There is no doubt thatSiddharšigaòi regarded NA.1 to aim at—beside enumerating subdivisions ofpramâòa (pratyakšaô ca parokšaô ca dvidhâ … )—formulating its definition(pramâòaô sva-parâbhâsi jñânaô bâdha-vivarjitam). And, further, there can hardlybe any doubt that a descriptive definition of pramâòa is precisely what Siddhasenain NA.1ab does. In the aphorism, he does not only group all cognitively validprocedures under respective categories of ‘direct’ and ‘indirect’, which is a case of atypological definition that demarcates a notion by mentioning specimens to whichthe definition is applicable (Ð di¦ tîn genîn ÐrismÕj), but he also attempts a realdefinition that describes the character and essential aspects or intrinsiccharacteristics of a thing (Ð lšgwn ™x ïn œstin ™nuparcÒntwn [ÐrismÕj]).37

Clearly, these two definitions represent two diverse approaches and the latter isintellectually more sublime. Whereas we quite frequently find typologicaldefinitions of pramâòa (viz. statements of its divisions) in Jaina literature, we donot, as a matter of fact, come across any attempt at a descriptive definition ofpramâòa of the second type in any works, both Jaina and non-Jaina, prior NA. Thus,the formulation of a descriptive definition of cognitive criterion seems definitely tobe an advancement. And it would perhaps be surprising to find such an innovationin a work of generally secondary character that repeatedly borrows from otherworks (it relies, for the most part, on the Buddhist legacy and has very little newideas, genuinely of its own, to offer), if NA. had no predecessor. However, it isDharmakîrti who must have inspired Siddhasena’s descriptive definition. Theopening verse (if we skip the first two introductory verses of obeisant nature) of thePramâòa-siddhi chapter, i.e. PV.1.3 and PV.1.7cd present such a descriptivedefinition of pramâòa that specifies its character and individual features. This

37 Cf. Aristotelian twofold division in Met.998b (p. 47.12–14): ›teroj d' œdtai Ð di¦

tîn genîn ÐrismÕj kaˆ Ð lšgwn ™x ïn œstin ™nuparcÒntwn.

ON THE DATE OF THE NYÂYÂVATÂRA 343 section of PV. is traditionally38 assumed to offer the first comprehensive definitionof pramâòa. Clearly, for both Manorathanandin and Siddharšigaòi the passages ofNA.1ab and PV.1.3/7cd respectively are cases of a pramâòa-lakšaòa.Occasionally39, a doubt is raised whether Dharmakîrti’s design was indeed adescriptive definition and that actually it was the subsequent Buddhist tradition ofcommentators where the idea of a comprehensive descriptive definition developed.Irrespective of whether one assumes that the formulation of PV.1.3 was a genuinedescriptive definition or that an idea of such a definition first originated with post-Dharmakîrtian commentators, both sides agree that such a descriptive definition didnot exist before Dharmakîrti (ergo it is altogether absent from Diónâga,Œaókarasvâmin, etc.). Accordingly, since NA. appears to present such a definition ofa descriptive character, it must have been composed after Dharmakîrti and PV.40

[4] Late Buddhist sources allegedly refer to NA.2.41 Surprisingly, NA.2, being astatement of pûrva-pakša, is the least suitable verse of the whole Dvâtriôœikâ toserve as an object of anybody’s critique. Thus, anyone referring to the verse with thepurpose to refute it would have to be highly incompetent and incapable of noticingthat the very verse is refuted by Siddhasena himself in NA.3! Therefore it could

38 See: FRANCO (1997: 59–61), esp.: ‘Therefore, we can safely conclude that the

definition of pramâòa in general, unlike the definition of each pramâòa in particular,was not undertaken by any Indian philosopher before Dharmakîrti’s time. Consequently,there is no reason why one should expect Dharmakîrti to do so. During the time thatseparates Dharmakîrti from his commentators, some change in the philosophicalrequirements must have occurred that produced the expectation of a general definition ofpramâòas.’ [p. 60]. Cf. also STEINKELLNER–KRASSER (1989: 3–5). Cf., e.g. PVV.1.3(1: 2.17 ff.; 3: 3.17 ff.): ayam âcâryo båhad-âcâryîya-pramâòa-samuccaya-œâstrevârttikaô cikîršuÿ svataÿ-kåta-bhagavan-namaskâraÿ tac-châstrârambha-samaye tad-âcârya-kåta-bhagavan-namaskâra-œlokaô vyâkhyâtu-kâmaÿ prathamaô pramâòa-sâmânya-lakšaòam âha … and PVV.1.7–8 (1: 4.36 ff.; 3: 8.6 ff.): tad evamavisaôvâdanaô pramâòa-lakšaòam uktam. idânîm anyad âha …

39 See: FRANCO (1997: 54–62), e.g.: ‘…Dharmakîrti did not attempt a generaldefinition of the pramâòas, but only wanted to prove that the Buddha is apramâòa … ’ [p. 61].

40 I am indebted to Claus Oetke for drawing my attention to the above issue ofPV.1.3–7 and its relevance to the problem of dating of NA.

41 QVARNSTRÖM (1999: 178): ‘In the eighth-century commentary on Dharmakîrti’sPramâòavârttika (II.5), Œâkyabuddhi (or Œâkyamati) quotes the second verse ofNyâyâvatâra and claims that this verse is the object of Dharmakîrti’s critique’, and refers top. 163, n. 38 of Chr. LINDTNER’s ‘Marginalia to Dharmakîrti’s Pramâòaviniœcaya.’ WienerZeitschrift für die Kunde Südasiens 28 (1984) 149–175.

344 PIOTR BALCEROWICZ hardly be an ‘object of Dharmakîrti’s critique.’ Furthermore, any reference toallegedly such a state of affairs would prove the source of such an information to bea highly unreliable one. Consequently, either Œâkyabuddhi cannot have referred toNA.2 or his acquaintance with Jaina tradition would be liable to doubt.

As a matter of fact, NA.2 may plausibly be taken as a prima facie objection (‘nopurpose is known for stating the definition of cognitive criterion’) againstformulating a comprehensive definition of pramâòa of descriptive character,inasmuch as ‘cognitive criteria (mind the plural!—P.B.) are well-known andeveryday practice is accomplished by them.’ Siddharšigaòi takes ‘the definition ofcognitive criterion’ to be ‘[a statement] consisting in mentioning specific propertiesof cognitive criterion that are capable of distinguishing [it from that which has]another form,’42 which is a very accurate explanation of a descriptive definition.Interestingly enough, the reason adduced why such a definition is not necessary incase of pramâòa is prasiddhâni pramâòâni, where the plural is expressly used. Thiswould point to the more archaic kind of definition based on classification, thatwould define the notion of pramâòa merely by specifying the pramâòa ‘inventory’.Thereby NA.2 supports our supposition that what Siddhasena really does is adescriptive definition of pramâòa. Accordingly, NA.2 would rather be a record ofthe historical change and certain resistance against a new approach towards definingpramâòa. This would also explain the role of the verse, that might seemunnecessary at first glance, in such a succinct work as NA. is.

[5] The pâda c (tad-vyâmoha-nivåttiÿ) of NA.343—which is a rejoinder to thedoubt raised in NA.2 whether it is ‘purposeful to state the definition of cognitivecriterion’ (viz. whether the treatise, œâstra = NA., has a purpose), since ‘cognitivecriteria are well-known and everyday practice is accomplished by them’—recallsDharmakîrti’s statement found in PV.1.7:

prâmâòyaô vyavahâreòa œâstraô moha-nivartanam /ajñâtârtha-prakâœo vâ svarûpâdhigateÿ param //

A statement expressing a similar idea seems to be absent from PS.

[6] Corresponding to the sequence of topics discussed in NB. (truly, it is a typicalsequence not restricted to NB.), as a next step, Siddhasena in NA.5 makes hispolemical statement and declares also inference to be ‘non-erroneous because it is a

42 NAV.2 ad loc.: pramâòa-lakšaòasyôktau para-rûpa-vyâvartana-

kšamâsâdhâraòa-pramâòa-dharma-kathana-rûpâyâô .43 prasiddhânâô pramâòânâô lakšaòôktau prayojanam /

tad-vyâmoha-nivåttiÿ syâd vyâmûðha-manasâm iha //

ON THE DATE OF THE NYÂYÂVATÂRA 345 cognitive criterion, just like perception.’ (… anumânaô, tad abhrântaô pramâòatvâtsamakšavat …). The use of the term abhrânta as well as the polemical characterdirectly points to NB.(1).1.4: tatra pratyakša ô kalpanâpoðham abhrântam. This issue,that was taken up by JACOBI and VAIDYA, has been already discussed above p. 333 ff.Dharmakîrti expressed such ideas that could have served both as a target andinspiration for Siddhasena also in PV.2.45–46 (anumâna-vicâraÿ)44.

But this is not the only Dharmakîrtian trace to be found in NA.5. Inference isdefined by Siddhasena to ‘determine the sâdhya on account of lióga; lióga (theinferential sign) is, in its turn, inseparably connected with the sâdhya; thedetermining factor is here the relation of avinâ-bhâva: NA.5ac: sâdhyâvinâ-bhunoliógât sâdhya-niœcâyakaô småtam / anumânaô. This definition follows, in most—ifnot all—details, the ideas expressed by Dharmakîrti in PV.1.287:

anumânâœrayo liógam avinâ-bhâva-lakšaòam /vyâpti-pradarœanâd dhetoÿ sâdhyenôktañ ca tat sphuþam //

This striking correspondence does not only concern central elements in inferenceand their character as well as their mutual connection, but even the choice ofvocabulary. The inseparable connection with the probandum as the definingcharacteristic of the probans reoccurs in NA.13: sâdhyâvinâ-bhuvo hetor.45

[7] NA.8 may have been influenced by Diónâga both in terms of vocabulary andnotions. Its pâda d (mânaô œâbdaô prakîrtitam) reminds of PS.(2).5.1 (esp. œâbda,sgra las byuó ba):

na pramâòântaraô œâbdam anumânât tathâ hi tat /kåtakatvâdivat svârtham anyâpohena bhâšate // 46

[8] In NA.13 Siddhasena takes a closer look at parârthânumâna and the conditionsof its validity: sâdhyâvinâ-bhuvo hetor vaco yat pratipâdakam / par ârtham anumâna ô tat pakšâdi-vacanâtmakam // The phrasing is reminiscent of NB.3.1: tri-rûpa-liógâkhyânaô par ârth ânumânam . Clearly, the element of tri-rûpa—a criterion ofvalidity of anumâna for the Dharmakîrtian tradition—is absent in NA., for this idea

44 ayathâbhiniveœena dvitîyâ bhrântir išyate / gatiœ cet para-rûpeòa na ca bhrânteÿ

pramâòâtâ // abhiprâyâvisaôvâdâd api bhrânteÿ pramâòatâ / gatir apy anyathâ dåšþâ,pakšaœ câyaô kåtôttaraÿ //

45 Cf. also PV.3.31: kârya-kâraòa-bhâvâd vâ svabhâvâd vâ niyâmakât / avinâ-bhâva-niyamo ’darœanân na na darœanât // , etc.

46 sgra las byuó ba rjes dpag las | tshad ma gŸan min de ltar de | byas sogs pa bŸindu raó don la | gŸan sel bas ni rjod par byed | The verse is quoted in TSaP. ad TSa.1514(p. 441.6–7), with a minor alteration (anumânât tathâ hi saÿ (tat) / ).

346 PIOTR BALCEROWICZ was not recognised by the Jainas. Instead, syntactically and semantically in the sameposition, we have the Jaina criterion of validity, viz. sâdhyâvinâ-bhuvaÿ. Thus, NA.13may be taken as Siddhasena’s polemical reply to the Buddhist doctrine of trairûpya.

On the other hand, pâdas c-d of NA.13 (…anumânaô tat pakšâdi-vacanâtmakam)bear close resemblance to (1) NP.(2).2 (= NP.(1), p. 1.4–5): tatra pakšâdi-vacanânisâdhanam. pakša-hetu-dåšþânta-vacanair hi prâœnikânâm apratîto ’rthaÿpratipâdyata iti // as well as to (2) the verse of NM.1: *pakšâdi-vacanânîtisâdhanam; tatra hi svayaô / sâdhyatvenêpsitaÿ pakšo viruddhârthânirâkåtaÿ // .However, this similarity is not decisive at all, inasmuch as all the quotations merelydescribe the widely known structure of any ‘syllogistic’ reasoning, that is composedof respective links, viz. members of ‘syllogism’.

[9] In the definition of pakša, NA.14ab: sâdhy âbhyupagamaÿ pakšaÿpratyakšâdy-anirâkåtaÿ / , we come across other formulations that bear obvioussimilarity especially to NB.(2).3.37: svarûpeò Îva svayam išþo ’nirâkåtaÿ pakša iti(but also to NB.(2).3.49–53). Dharmakîrti’s svarûpeòa stands for sâdhyatvena,which is confirmed both by Dharmakîrti himself in the next two sûtras(NB.(2).3.39–40: svarûpeòêti sâdhyatvenêšþaÿ. svarûpeòÎvêti sâdhyatvenÎvêšþo nasâdhyatvenâpi.) as well as by Dharmottara (NBÞ.(2).3.37 ad loc.: svarûpeòÎvêtisâdhyatvenÎva.). Dharmottara explains anirâkåtaÿ as pratyakšâdy-anirâkåtaÿ(NBÞ.(2).3.37 ad loc.: evaô-bhûtaÿ san pratyakšâdy-anirâkåto yo ’rthaÿ sa pakšaity ucyate.), and—if we suppose that he expressed original ideas of Dharmakîrti(and I believe he did in this regard)—Siddhasena’s formulations express almost thesame idea as Dharmakîrti did in NB., barring Dharmakîrti’s specific delimiting useof eva as well as svayam.

However, there can hardly be any doubt that the formulations of NA.14ab:sâdhy âbhyupagamaÿ pakšaÿ pratyakšâdy-anirâkåtaÿ / , go back to PV.4.86(1: p. 378, 3: p. 390):

sâdhy âbhyupagama ÿ pakša-lakšaòaô tešv apakšatâ /nirâkåte47 bâdhanataÿ œeše ’lakšaòa-våttitaÿ //

The idea itself goes back to Diónâga and NM. (*svayaô sâdhyatven êpsitaÿ pakšoviruddhârthânirâkåtaÿ), as it was pointed out by Manorathanandin inPVV., p. 378.26. Another plausible source for NA.14 might be PS.3.2:

svarupeòÎva nirdeœyaÿ svayam išþo ’nirâkåtaÿ /pratyakšârthânumânâpta-prasiddhena svadharmiòi // 48

47 Cf. Manorathanandin’s Våtti ad loc.: tathâ ca tešu œâstrêšþâdišu pañcasu

vyâvartyešu mâdhye nirâkåte pratyâkšâdi-bâdhite bâdhanato ’pakšatâ viruddhârthâ.

ON THE DATE OF THE NYÂYÂVATÂRA 347 Clearly, the word âdi of the phrase pratyakšâdi of NA.14b might allude to thecategories [pratyakšârtha?], anumâna, âpta, prasiddha enumerated in Diónâga’sanirâkåtaÿ pratyakšârthânumânâpta-prasiddhena, but it does not necessarily haveto, since Siddharšigaòi (NAV.14 ad loc.: âdi-œabdâd anumâna-sva-vacana-lokâgåhyante) takes âdi to stand for anumâna-sva-vacana-loka. Another predecessor ofDharmakîrti in this regard was Œaókarasvâmin49.

Out of all these possible sources, NA.14 has most in common with Dharmakîrti interms of (1) exact wording (sâdhyâbhyupagamaÿ pakša … ), (2) replacement ofDiónâgan išþa with abhyupagama, (3) affinity in the explicit correlative sâdhya(instead of Diónâgan svarupeòÎva nirdeœya) to abhyupagama / išþa.

[10] The description of hetu and its role in the inference for others(parârthânumâna) in NA.17 is not so conspicuously similar to NB. in phrasing.Nevertheless, semantically both expositions are quite akin to each other. That inNA.17 we still deal with parârthânumâna is clear from the context itself, but alsoSiddharšigaòi leaves not doubt (NAV.17 ad loc.: parârthânumânasya vacana-rûpatvâd … ). Thus, in both cases we have ‘the pronouncement of the logicalreason’ (hetos … prayogo, NA.) or ‘the announcement of the inferential sign’(liógâkhyânaô, NB.) as the principal element of the inference for others(parârthânumâna) and the idea that there is no difference in ‘demonstrative force’between the two formulations of the logical reason:

NA.17: hetos tathôpapattyâ vâ syât prayogo ’nyathâpi vâ /dvi-vidho ’nyatareò âpi sâdhya-siddhir bhaved iti //

NB.3.1,3–7: /1/ tri-rûpa-liógâkhyânaô parârthânumânam. /3/ taddvi-vidham. /4/ prayoga-bhedât. /5/ sâdharmya-vaidharmyavac cêti./6/ n ânayor arthataÿ kaœcid bhedaÿ . /7/ anyatra prayoga-bhedât .

The idea expressed by Dharmakîrti goes back to his PV.3.16 and PVSV. ad loc.50

48 The above Sanskrit reconstruction of PS.3.2 follows FRAUWALLNER (1957/b: 885)

and TILLEMANS (1997: 178, n. 2). Tibetan text reads as follows: raó gi óo bo kho naa

bstan | bdag ’dod raó gi chos can la | móon sum don daó rjes dpag daó | yid ches gragspas ma bsal ba’o || [a TILLEMANS 1997 reads nar.] PS.3.2 must have been in its turn thesource for NB.(2).3.37.

49 NP.(2).2.1: tatra pakšaÿ prasiddho dharmî prasiddha-viœešaòa-viœišþatayâ svayaôsâdhyatven êpsitaÿ . pratyakšâdy-viruddha iti vâkya-œešaÿ.

50 In PV.3.15 [p. 180] Dharmakîrti first recalls the triple character of hetu defined byDiónâga in Nyâya-mukha: hetos trišv api rûpešu niœcayas tena varòitaÿ / asiddha-viparîtârtha-vyabhicâri-vipakšataÿ // . Then (PV.3.16ab) he states the reason for


[11] In NA.20 Siddhasena maintains that dåšþânta is not an essential part of‘syllogistic’ reasoning, inasmuch as the relation of invariable concomitance (vyâpti)suffices to prove the thesis. This is a continuation of the ‘economical’ trend inIndian logic—that starts with Vasubandhu and his Vâda-vidhâna and Vâda-vidhi51—to limit the number of necessary ‘syllogistic’ members, to simplify thereasoning procedures and to make such procedures universally binding, without anyneed for further empirical justification than the premises themselves:

NA.20: antar-vyâptyÎva sâdhyasya siddher bahir-udâhåtiÿ /vyarthâ syât tad-asadbhâve ’py evaô nyâya-vido viduÿ //

NB.(2).3.121: tri-rûpo hetur uktaÿ. tâvatâ cârtha-pratîtir iti na påthag dåšþ ânto nâma sâdhan âvayavaÿ kaœcit . tena nâsya lakšaòaô påthagucyate gatârthatvât.

Siddhasena was not so much innovative as it might seem at first glance, inasmuchas it is Dharmakîrti (NB.(2).3.121) who had intuitively foreshadowed such an ideabefore him, when he had claimed that the example (dåšþânta) is not a separatemember of the proof (sâdhanâvayava). It is much more natural and less surprisingto find the direct continuation of the ‘economical’ trend, that had started withVasubandhu, in the Yogâcâra tradition and Dharmakîrti’s works, rather than in theJaina tradition in the person of Siddhasena, where we do not find such an‘economic’ tendency before Siddhasena. In my opinion, Siddhasena verbalised whathad already been latent in Dharmakîrti’s statement, even though Dharmakîrtihimself had not been able to do without the example, which he had considered anintegral part of the logical reason (hetu), indispensable to authenticate the generalprinciple by taking recourse to its instantiations (relevant portions underlined):NB.(2).3.122: hetoÿ sapakša eva sattvam asapakšâc ca sarvato vyâvartî rûpamuktam abhedena. punar viœešeòa kârya-svabhâvayor ukta-lakšaòayor janma-tan-mâtrânubandhau darœanîyâv uktau. tac ca darœayatâ—yatra dhûmas tatrâgnir,

expressing the concomitance by way of dissimilarity (vaidharmya-vacana): vyabhicâri-vipakšeòa vaidharmya-vacanaô ca yat / , only to quote Diónâga in PVSV.(1).3.16ab: yadâha—eša tâvan nyâyo yad ubhayaô vaktavyaô viruddhânaikântika-pratipakšeòa iti. Therule is further explained by Dharmakîrti in PVSV. as follows: sâdharmya-vâcanaôviruddha-pratipakšeòa, vaidharmya-vâcanam anaikântika-pratipakšeòa. Thereupon, inPV.3.16cd and PVSV. ad loc., he adds that: yady adåšti-phalaô tac ca yadi tena vipakše’darœanaô khyâpyate tad anukte ’pi gamyate // . Cf. also PVSV.(1).3.24–25 [p. 185]:tasmât svabhâva-pratibandhâd eva hetuÿ sâdhyaô gamayati. sa ca tad-bhâva-lakšaòastad-utpatti-lakšaòo vâ. sa evâvinâ-bhâvo dåšþântâbhyâô pradarœyate.

51 Cf. FRAUWALLNER (1933) and FRAUWALLNER (1957/a).

ON THE DATE OF THE NYÂYÂVATÂRA 349 asaty agnau na kvacid dhûmo yathâ mahânasêtarayor, yatra kåtakatvaôtatrânityatvam, anityatvâbhâvo kåtakatvâsambhavo yathâ ghaþâkâœayor—itidarœanîyam. na hy anyathâ sapakša-vipakšayoÿ sad-asattve yathôkta-prakâre œakyedarœayitum. tat-kâryatâ-niyamaÿ kârya-liógasya, svabhâva-liógasya ca svabhâvenavyâptiÿ. asmiôœ cârthe darœite eva dåšþânto bhavati. etâvan-mâtra-rûpatvât tasyêti.Dharmakîrti expressed a similar idea already in his PV.3.27: tad-bhâva-hetu-bhâvauhi dåšþânte tad-avedinaÿ / khyâpyete, vidušâô vâcyo hetur eva hi kevalaÿ // .Siddhasena, however, has to his credit that he clearly states the conditions ofinternal formal validity of the proof and dismisses the need to quote anyinstantiation: the proof is valid because the premisses are valid and the relationbetween them is universally binding.

[12] Siddhasena states explicitly five conditions that invalidate pakša—and addsthat there are, in fact, several varieties of pakšâbhâsa—in NA.21:

pratipâdyasya yaÿ siddhaÿ pakšâbhâso ’kša-liógataÿ /loka-sva-vacanâbhyâô ca bâdhito ’nekadhâ mataÿ //

These five pakšâbhâsas are in concord with the list exemplified by Siddharšigaòi inNAV.: (1) pratipâdya-siddha, (2) pratyakša-bâdhita, (3) anumâna-bâdhita,(4) loka-bâdhita, (5) sva-vacana-bâdhita. Having enumerated five varieties ofpakšâbhâsa, it would be redundant and pointless on Siddhasena’s part to say by wayof recapitulation in a succinct kârikâ that these varieties are numerous: anekadhâmataÿ, unless he had other varieties, not mentioned already by name, in mind. Thatbeing the case, he probably referred to other enumerations well-known from othersources.

In fact, Siddhasena’s enumeration overlaps with Dharmakîrti’s list of fallacies ofthe thesis, enumerated in NB.(2).49–5352. Interestingly, Dharmakîrti subsequently(NB.(2).3.5453) adds a few more conditions and the failure to meet them wouldrender the pakša defective as well. Thus, Siddhasena—by anekadhâ mataÿ—may

52 /49/ (2) tatra pratyakša-nirâkåto yathâ: aœrâvaòaÿ œabda iti. /50/ (3) anumâna-

nirâkåto yathâ: nityaÿ œabda iti. /51/ (4?) pratîti-nirâkåto yathâ: acandraÿ œaœîti. /52/ (5)sva-vacana-nirâkåto yathâ: nânumânaô pramâòam. /53/ iti catvâraÿ pakšâbhâsânirâkåtâ bhavanti.

53 evaô siddhasya, asiddhasyâpi sâdhanatvenâbhimatasya, svayaô vâdinâ tadâsâdhayitum anišþasya, ukta-mâtrasya nirâkåtasya ca viparyayeòa sâdhyaÿ. tenÎvasvarûpeòâbhimato vâdina išþo ’nirâkåtaÿ pakša iti pakša-lakšaòam anavadyaôdarœitaô bhavati.

350 PIOTR BALCEROWICZ have referred to Dharmakîrti’s catalogue of defective pakšas. But not necessarily toDharmakîrti’s. He may have as well referred, e.g. to NP.(2).3.154.

In any case, NA.21 can in no way attest that Siddhasena was posterior or prior toDharmakîrti. It merely points to certain similarities between Siddhasena’s list andthe Buddhist tradition. Probably, it was the tradition of Œaókarasvâmin andDharmakîrti that Siddhasena alluded to by anekadhâ mataÿ.

[13] In Siddhasena’s subsequent aphorism, we find further similarities with NB.,though they are less of linguistic nature (similarities in formulations) but rather ofmethodological character. Analogously to the structure of NB, Siddhasena—afterdescribing fallacious theses—proceeds to discuss fallacious logical reasons (hetv-âbhâsa). However, both the authors first look back to their previous definitions of acorrect, not defective hetu:

NA.22: anyathânupapannatvaô hetor lakšaòam îritam55 /tad-apratîti-sandeha -viparyâsais tad-âbhatâ //

NB.(2).3.55–56: tri-rûpa-liógâkhyânaô parârthânumânam ityuktam56. tatra trayâòâô rûpâòâm ekasyâpi rûpasyânuktausâdhanâbhâsaÿ. uktâv apy asiddhau sandehe vâ pratipâdya-pratipâdakayoÿ.

Incidentally, it is the only case in both works that the authors first remind thereader / hearer of the definition of a correct ‘syllogistic’ member / term, and only thendeal with its particular fallacies.

Likewise incidentally, both the authors first state general factors that invalidate acorrect hetu, either singly or jointly, and subsequently enumerate resulting fallaciesone by one. There are two such general invalidating factors for Dharmakîrti, viz.asiddhi and sandeha (NB.3.56,109): various combinations (with regard totrairûpya) are responsible for particular varieties of hetv-âbhâsa. For Siddhasena,however, there are three—polemically, as it were—such factors: tad-apratîti,sandeha and viparyâsa; thus Dharmakîrti’s asiddhi factor would seem to bifurcate

54 sâdhayitum išþo ’pi pratyakšâdi-viruddhaÿ pakšâbhâsaÿ, tad yathâ: (1) pratyakša-

viruddhaÿ, (2) anumâna-viruddhaÿ, (3) âgama-viruddhaÿ, (4) loka-viruddhaÿ, (5) sva-vacana-viruddhaÿ, (6) aprasiddha-viœešaòaÿ, (7) aprasiddha-viœešyaÿ,(8) aprasiddhôbhayaÿ, (9) prasiddha-sambandhaœ cêti // tatra …

55 Either in NA.5 (according to Siddharši) or in NA.17 (hetos tathôpapattyâ vâ syâtprayogo ’nyathâpi vâ, see NAV. ad loc.: anyathâpi vêty anenâvayavesamudâyôpacârâd anyathânupapattiô lakšayati).

56 NB.3.1: tri-rûpa-liógâkhyânaô parârthânumânam; cf. also NB.(2).2.5,11–12.

ON THE DATE OF THE NYÂYÂVATÂRA 351 into Siddhasena’s tad-apratîti and viparyâsa. For instance, for Siddhasena thefallacies of NB.3.58 and 5957 would be probably a case of viparyâsa, whereas thefallacy of NB.3.6058 should rather be considered a case of tad-apratîti. One couldanalyse all the remaining cases of fallacious hetus (ensuing from the combination ofthe factors asiddhi and sandeha) found in NB. and map them onto the tripleclassification of Siddhasena in the same manner. Practically, it would mean todecide which of the asiddhi cases of Dharmakîrti would correspond to tad-apratîticases of Siddhasena, and which to his viparyâsa cases. However, we do not haveany explicit statement in the far too succinct NA., nor in NAV., that would provideus any algorithm of such a mapping, and the issue is open to our conjectures only.

Interestingly, in NP., for instance, to which NA. might be thought to haveoccasionally referred to, we do not find any trait of such invalidating factors asasiddhi and sandeha, or anything similar.

[14] Also NA.23 and the classification of particular varieties of hetv-âbhâsaspoints to secondary sources of Siddhasena’s ideas. There seem to be only two suchpotential sources, i.e. NB. and NP.:

NA.23: asiddhas tv apratîto yo yo ’nyathÎvôpapadyate /viruddho yo ’nyathâpy atra yukto ’naikântikaÿ sa tu //

NB.(2).3.109: evam ešâô trayâòâô rûpâòâm ekÎkasya dvayor dvayorvâ rûpayor asiddhau sandehe vâ yathâ-yogam asiddha-viruddhânaikântikâs trayo hetv-âbhâsâÿ.

NP.(2).3.2. asiddhânaikântika-viruddhâ hetv-âbhâsâÿ //

However, NB. is a more probable source, inasmuch as the sequence of fallacies(asiddha, viruddha, anaikântika) listed in NA. is exactly the same as that of NB.,not of NP., where the two last varieties are interchanged (asiddha, anaikântika,viruddha). Diónâga has the aniœcita (or sandigdha) variety, istead of anaikântika.

What is striking is that Siddhasena has only three varieties of hetv-âbhâsa, likeDharmakîrti and Œaókarasvâmin, unlike Diónâga. Moreover, Dharmakîrti

57 NB.(2).3.57–9: /57/ ekasya rûpasya dharmi-sambandhasyâsiddhau sandehe

vâsiddho hetv-âbhâsaÿ. /58/ yathâ: anityaÿ œabda iti sâdhye câkšušatvamubhayâsiddham [not proved for both parties]. /59/ cetanâs tarava iti sâdhye sarva-tvag-apaharaòe maraòaô prativâdy-asiddham, vijñânêndriyâyur-nirodha-lakšaòasyamaraòasyânenâbhyupagamât, tasya ca tarušv asambhavât [not proved for theopponent].

58 NB.(2).3.60: acetanâÿ sukhâdaya iti sâdhya utpattimattvam anityatvaô vâsâôkhyasya svayaô vâdino ’siddham [not proved for the proponent himself].

352 PIOTR BALCEROWICZ outspokenly rejects Diónâga’s subvariety, i.e. viruddhâvyabhicârin59: NB.(2).3.110:viruddhâvyabhicâry api saôœaya-hetur uktaÿ. sa iha kasmân nôktaÿ. We do notfind any polemical trait in NA., which could even suggest that Siddhasena adoptedDharmakîrti’s criticism of viruddhâvyabhicârin. Apparently he subscribed to thecriticism unhesitatingly.

It is worth noticing that NA. diverges also from the tradition of Kumârila, whosetriple classification—into asiddha, sandigdha, viruddha in MŒV.—is more akin toDiónâga’s: MŒV.5.4.75 p. 264 mentions the classification (saôœayâdî-viparyâyâÿ);subsequently (1) the threefold asiddha fallacious reason is mentioned inMŒV.5.4.76–83ab; (2) sandigdha and (3) viruddha occur in MŒV.5.4.83cd–107(e.g. in MŒV.5.4.83cd: sandeha-viparîtatva-hetû câtra nirâkåtau); besides,(2) sandigdha is further found in MŒV.5.4.84b–96b, whereas (3) viruddha ismentioned in MŒV.5.4.96cd–107ab. Kumârila nowhere in the Anumâna-paricchedasection uses the term anaikântika as a hetv-âbhâsa, he has sandigdha instead, likeDiónâga, unlike Dharmakîrti, Œaókarasvâmin and Siddhasena. However,Pârthasârathi Miœra follows the general post-Dharmakîrtian typology in hisclassification, while commenting on Kumârila (p. 264.11 ad MŒV.5.4.75): sampratihetv-âbhâsân asiddhânaikântika-viruddhân prapañcayan …

[15] NA.24–25 reveal further similarities pointing to NB. as its possibleinspiration. One of them is the phraseological affinity as regards the use of dåšþânta-doša, instead of dåšþântâbhâsa, in both works. Both Siddhasena and Dharmakîrtiuse the expression pakšâbhâsa as well as similarly tad-âbha and hetv-âbhâsa.However, both of them deviate from the general use of derivatives of âÖbhâs totechnically denote logical fallacies, when they refer to dåšþânta by the term doša,and to dåšþânta only:

NA.24: sâdharmyeòâtra dåšþânta-došâ … ,NA.25: vaidharmyeòâtra dåšþânta-došâ … ,

NB.(2).3.123, 128–129: /123/ etenÎva dåšþânta-došâ api nirastâbhavanti. /128/ sâdharmyeòa dåšþânta-došâÿ. /129/ vaidharmyeòâpi:paramâòuvat karmavad âkâœavad iti sâdhyâdy-vyatirekiòaÿ.

Indeed, also Œaókarasvâmin has dåšþânta-doša twice, but in a slightly differentcontext, viz. that of refutation. Generally, when he discusses fallacious examples, heuses the standard term dåšþântâbhâsa, e.g. NP.3.3: dåšþântâbhâso dvividhaÿ:

59 Cf. also RANDLE (1926: 68–69, 79).

ON THE DATE OF THE NYÂYÂVATÂRA 353 sâdharmyeòa vaidharmyeòa ca // The only two occurences of dåšþânta-doša arefound in NP.(2).6,760.

Conspicuously, dåšþânta-doša occurs as exemplification of dûšaòâni, in the seriesof sâdhana-doša, pakša-doša, hetu-doša and dåšþânta-doša.61 Another occurrence ofdoša (however, pratijñâ-doša, not dåšþânta-doša) in NP. is attested in theconcluding lines of NP.(2).3.1, that summarise the discussion on pakšâbhâsa62. Alsohere, the term doša has a more general meaning than the technical term âbhâsa, andthe peculiarity of the usage of pratijñâ-doša is confirmed by the closing uktâÿpakšâbhâsâÿ. The usage of doša in NP. confirms the fact that in works precedingDiónâga, Dharmakîrti or Œaókarasvâmin—e.g. in NS., VS., NBh., PBh., etc.—došais employed to denote general defects and is not used specifically as a terminustechnicus in the sense of âbhâsa.

[16] There is a structural similarity to be observed: both Siddhasena (NA.24) andDharmakîrti (NB.(2).3.122–3) explicitly define fallacious examples by referring tothe definition of a correct hetu. Accordingly, NB.3.122 recapitulates the definitionof correct hetu as well as conditions of its validity. The successive statement ofNB.3.123 (etenÎva dåšþânta-došâ api nirastâ bhavanti; vide supra p. 352) showsthat fallacious examples are refuted by referring to the correct definition of hetu, andall fallacies of the example share the same characteristic. Similarly, Siddhasenarelates the deficiency of dåšþântâbhâsas to deficient logical reasons and, withapalakšaòa-hetûtthâÿ, he expresses the idea known from the work of Dharmakîrti.

[17] Strangely, in NA.24–25 Siddhasena refers to some tradition by nyâya-vid-îritâÿas regards the classification of fallacious examples. However, there seems to havebeen no earlier (or contemporary) Jaina source he could have referred to. Thus, indefault of any extant evidence attesting to a Jaina tradition which offered a typology of

60 (6) sâdhana-došôdbhâvanâni dûšaòâni // sâdhana-došo nyûnatvam. pakša-došaÿ

pratyakšâdi-viruddhatvam. hetu-došo ’siddhânaikântika-viruddhatvam. dåšþânta-došaÿsâdhana-dharmâdy-asiddhatvam. tasyôdbhâvanaô prâœnika-pratyâyanaô dûšaòam //(7) abhûta-sâdhana-došôdbhâvanâni dûšaòâbhâsâni // saôpûròe sâdhane nyûnatva-vacanam. adušþa-pakše pakša-doša-vacanam. siddha-hetuke ’siddha-hetukaô vacanam.ekânta-hetuke ’nekânta-hetukaô vacanam. aviruddha-hetuke viruddha-hetukaôvacanam. adušþa-dåšþânte dušþa-dåšþânta-doša-vacanam. etâni dûšaòâbhâsâni. na hyebhiÿ para-pakšo dûšyate, niravadyatvât tasya // ity uparamyate // .

61 NP.6: sâdhana-došôdbhâvanâni dûšaòâni // sâdhana-došo nyûnatvam. pakša-došaÿpratyakšâdi-viruddhatvam. hetu-došo ’siddhânaikântika-viruddhatvam. dåšþânta-došaÿsâdhana-dharmâdy-asiddhatvam. tasyôdbhâvanaô prâœnika-pratyâyanaô dûšaòam // .

62 ešâô vacanâni dharma-svarûpa-nirâkaraòa-mukhena pratipâdanâsaôbhavataÿsâdhana-vaiphalyataœ cêti pratijñâ-došâÿ // uktâÿ pakšâbhâsâÿ //

354 PIOTR BALCEROWICZ fallacious examples, one is prone to assume—unless we find any indication to thecontrary—that, apparently, he must have referred to a general Indian tradition, wherewe do find such a typology of sâdharmya-° and vaidharmya-dåšþântâbhâsas.

Besides, he also seems to have taken it for granted that the reader / hearer couldeasily determine what is meant by sâdhyâdi-vikalâdayaÿ in NA.24. This task wouldnaturally be quite easy for anyone who was acquainted with NB.(2).3.124–125. Atthe same time, however, his elliptical formulation sâdhyâdi-vikalâdayaÿ was,polemically as it were, unequivocal enough to rule out the three remaining varieties,viz. (A7) ananvaya, (A8) apradarœitânvaya and (A9) viparîtânvaya, formulated byDharmakîrti (NB.(2).3.126–127), which could by no means have been hinted at bythe formulations of NA.24.

Similarly, the sixfold classification of dissimilar dåšþântâbhâsas systematised inNA.25 (sâdhya-sâdhana-yugmânâm anivåtteœ ca saôœayât) closely corresponds tothe first six fallacious examples of NB.(2).3.129–132: (V1) sâdhyâvyatirekin,(V2) sâdhanâvyatirekin, (V3) sâdhya-sâdhanâvyatirekin, (V4) sandigdha-sâdhya-vyatireka, (V5) sandigdha-sâdhana-vyatireka, (V6) sandigdha-sâdhya-sâdhana-vyatireka. Seemingly, NA. and NB. are at variance as regards terminology, forSiddhasena’s formulation: sâdhya-sâdhana-yugmânâm anivåtteœ ca saôœayât,differs from Dharmakîrti’s (V1)–(V6). In my opinion, however, anivåtteœ andsaôœayât of NA.25 indicate rather plainly °−vyâvåtta and sandigdha−° as the last andfirst elements of the compounds (V1)–(V3) and (V4)–(V6), respectively.Optionally, we could have (V1) *sâdhyânivåtta, (V2) *sâdhanânivåtta,(V3) *sâdhya-sâdhanânivåtta, (V4) *saôœayita-sâdhya, (V5) *saôœayita-sâdhana,(V6) *saôœayita-sâdhya-sâdhana, which is not different at all from the ideaexpressed in NB.(2).3.129–132.63

63 True, theoretically speaking, one could also interpret NA.25 to enforce the

acceptance of only (1) sâdhyâvyatirekin, (2) sâdhanâvyatirekin, (3) sâdhya-sâdhanâvyatirekin—to use Dharmakîrti’s terminology—and only one or more varietiesout of (4) sandigdha-sâdhya-vyatireka, (5) sandigdha-sâdhana-vyatireka,(6) sandigdha-sâdhya-sâdhana-vyatireka, but not necessarily all of them. Whileanivåtteÿ is unquestionably construed with sâdhya-sâdhana-yugmânâm, the expressionsaôœayât might be conjectured to be taken separately, without any dependence on allelements of sâdhya-sâdhana-yugmânâm. In this respect, attention should be drawn to aninstance of different conceivable ways of construing an aphorism of NA. bycommentators which we encounter in the case of NA.8: dåšþêšþâvyâhatâd vâkyâtparamârthâbhidhâyinaÿ / tattva-grâhitayôtpannaô mânaô œâbdaô prakîrtitam // Theaphorism is explained differently by the commentators, viz. (1) NAV. ad loc.: dåšþenapramâòâvalokitenêšþaÿ pratipâdayišito ’vyâhato ’nirâkåtaÿ sâmarthyâd artho yasminvâkye tat-tathâ; and (2) NAÞ. ad loc.: dåšþenêty-âdi. ayaô bhinnâdhikaraòas tri-pado

ON THE DATE OF THE NYÂYÂVATÂRA 355

As for another possible source of insipration, also NP.(2).3.3 distinguished—inaccord with the prevalent tradition of those days—two general categories: similarand dissimilar fallacious examples (dåšþântâbhâso dvividhaÿ: sâdharmyeòavaidharmyeòa ca // ). However, a closer look at Œaókarasvâmin’s varieties revealsessential differences, apart from the terminological ones. Accordingly, I see no wayhow the formulations of fallacies of examples based on similarity found in Nyâya-praveœa64 could be interpreted to have influenced Siddhasena’s sâdhyâdi-vikalâdayaÿ. As regards Nyâya-praveœa on fallacies of examples based ondissimilarity65, the influence might have been restricted to (V1), (V2) and (V3) only.

[18] Also NA.26, where criticism / refutation (dûšaòa) and its fallacy(dûšaòâbhâsa) are discussed, might have partly been inspired by Dharmakîrti’sNB.(2).3.137–140. Thus, NA.26a-c: vâdy -ukte sâdhane prokta-došâòâmudbhâvanam / dûšaòaô, reveals a certain similarity—in terms of both formulationsand ideas—to the dûšaòa-section of NB.(2).3.137–8: /137/ dûšaòâ nyûnatâdy-uktiÿ./138/ ye pûrvaô nyûnatâdayaÿ sâdhana -došâ uktâs tešâm udbhâvanaô dûšaòaô.tena par êšþ ârtha-siddhi-pratibandhât. Less conspicuous, though, is the remainingportion of the second hemistich of NA.26: niravadye tu dûšaòâbhâsa-nâmakam //The corresponding dûšaòâbhâsa-section of NB.(2).3.139–140 reads:/139/ dûšaòâbhâsâs tu jâtayaÿ. /140/ abhûta-došôdbhâvanâni jâty-uttarâòîti. Thereoccurring element dûšaòâbhâsa is not decisive at all, whereas Siddhasena’sniravadye could be a vague echo of Dharmakîrti’s abhûta-doša.

bahu-vrîhiÿ yadi vâ išþo ’vyâhato ’rtho yatra tad išþâvyâhataô vâkyam, tadanu dåšþenapramâòa-niròîtena išþâvyâhatam iti tat-purušaÿ (cf. BALCEROWICZ (1999: 4, n. 8)).Definitely, such an unnatural interpretation—i.e. to take anivåtteœ to refer to all elementsof the triad sâdhya-sâdhana-yugmânâm, while limiting the scope of saôœayât toselected element(s) of the compound sâdhya-sâdhana-yugmânâm—would be a mereguesswork, and one would rather, as a rule, construe sâdhya-sâdhana-yugmânâm withboth anivåtteœ and saôœayât, and obtain six varieties of dissimilar dåšþântâbhâsas. Onewould not, in any case, obtain any further varieties mentioned by Dharmakîrti inNB.(2).3.133–135: (V7) avyatireka, (V8) apradarœita-vyatireka and (V9) viparîta-vyatireka. Thus, Siddhasena apparently does not accept without reservation the Buddhisttypology by rejecting (V7), (V8) and (V9).

64 NP.(2).3.3.1: tatra sâdharmyeòa tâvad dåšþântâbhâsaÿ pañca-prakâraÿ, tad yathâ:(1) sâdhana-dharmâsiddhaÿ, (2) sâdhya-dharmâsiddhaÿ, (3) ubhaya-dharmâsiddhaÿ,(4) ananvayaÿ, (5) viparîtânvayaœ cêti // tatra …

65 NP.(2).3.3.2. vaidharmyeòâpi dåšþântâbhâsaÿ pañca-prakâraÿ, tad yathâ:(1) sâdhyâvyâvåttaÿ, (2) sâdhanâvyâvåttaÿ, (3) ubhayâvyâvåttaÿ, (4) avyatirekaÿ,(5) viparîta-vyatirekaœ cêti // tatra …


As a matter of fact, NA.26c-d (niravadye tu dûšaòâbhâsa-nâmakam) betrays moreaffinity to the closing section of NP.(2).7: abhûta-sâdhana-došôdbhâvanânidûšaòâbhâsâni // … etâni dûšaòâbhâsâni. na hy ebhiÿ para-pakšo dûšyate,niravadyatvât tasya // ity uparamyate // It is this section, in all probability, thatinfluenced both Dharmakîrti’s NB.(2).3.139–140 (dûšaòâbhâsâs tu jâtayaÿ. abhûta-došôdbhâvanâni jâty-uttarâòîti.) and the portion of NA.26c-d in question.

[19] The twenty-sixth aphorism (the exposition of dûšaòa) is the last section ofthe Nyâyâvatâra, where possible influences from Dharmakîrti’s side—in terms ofSiddhasena’s direct use of Dharmakîrti vocabulary or his response to Dharmakîrti’sideas—are easily detectable. Strangely enough, the topic dealt with in NA.26closely corresponds to the final issue discussed by Dharmakîrti in NB. Thus, theconspicuous absence of further possible Dharmakîrtian traces in NA.—theoreticallyderivable from other works of Dharmakîrti—points, in my opinion, to the fact thatSiddhasena—while composing NA.—closely followed the structure and thecontents of NB., up to NA.26.

In the remaining aphorisms (28–32) Siddhasena discusses issues peculiar to Jainism(viz. corollaries of kevala-jñâna and syâd-vâda) and there could hardly have been anyBuddhist influence to be noticed in any case: [27] the character of absolute cognition(kevala-jñâna); [28] the result of valid cognition in general; the results of absolutecognition (kevala-jñâna); the results of valid cognitive procedures other than kevala-jñâna; [29] the multiplex character of reality; the domain of cognitive acts; the domainof viewpoints (naya); [30] the character of viewpoints (naya); the description of thedoctrine of seven-fold modal description (syâd-vâda); [31] the character of thecognitive subject, the soul (jîva); [32] the eternal character of Jaina epistemology.

[20] The phrase pramâòa-phala occurring in NA.28 is occasionally taken to be aproof of its dependence on Diónâga.66 It is commonly assumed that the phrase inquestion goes back to Diónâga, e.g. PS.(1).1.8cd–1067. In these verses Diónâga

66 Cf., e.g. QVARNSTRÖM (1999: 178): ‘Furthermore, the Nyâyâvatâra (28) uses the

signature element of Dignâga, namely “pramâòaphala”61;’ in his note 61, he furtherdraws the reader’s attention to the work of G. Dreyfus and Chr. Lindtner: ‘The YogâcâraPhilosophy of Dignâga and Dharmakîrti”. Studies in Central & East Asian Religions.Vol. 2, Ed. by Per K. Sørensen et al. Copenhagen 1989: 27–52.

67 savyâpâra-pratîtitvât pramâòaô phalam eva sat // sva-saôvittiÿ phalaô vâtra tad-rûpo hy artha-niœcayaÿ / višayâkâratÎvâsya pramâòam tena mîyate // yad-âbhâsaôprameyaô tat pramâòa-phalate punaÿ / grâhakâkâra-saôvittî trayaô nâtaÿ påthak-kåtam // . For the Sanskrit text, see HATTORI (1968: 97, n. 1.55—107, n. 1.67). Tibetantext reads as follows:

ON THE DATE OF THE NYÂYÂVATÂRA 357 asserts that the result of pramâòa is pramâòa itself, or introspective cognition whichconsists in the determining of an object (sva-saôvitti, artha-niœcaya). Precisely thesame idea (âtmânubhâva, artha-viniœcaya) is echoed in PV.2.306–307ab, 33968.Generally, the idea is discussed at length by Dharmakîrti both in hisPVin.I(1).78.12–100.26 and in PV.2.301–366, 388–391. Accordingly, NA.28 mightbe taken to be a rejoinder of both Diónâga and Dharmakîrti, and there seems tonothing decisive to be found in NA.28 that would exclude any of the two authors.

On the other hand, any attempt to look for inspiration of NA.28 in works of eitherDiónâga or Dharmakîrti seems to me to be a result of misapprehension of the trueimport of the aphorism. In it, Siddhasena does not discuss the problem whetherpramâòa-phala can or cannot be equated with pramâòa itself, which is the majorconcern of Diónâga and Dharmakîrti. What is intended in the verse are rather extra-epistemological issues of both soteriological (kevala) and mundane (œeša) character,viz. the result of pramâòa is ‘the cessation of nescience’ (ajñâna-vinivartana),whereas the result of specific kinds of cognition is two-fold. The first categorysubsumes happiness and indifference (sukhôpekše) in case of the perfect knowledge(kevala), being a prerequisite of liberation (mokša) and commonly taken by the Jainasto be tantamount to the destruction of nescience. Since it results from the destructionof karman69 it is necessarily associated with innate happiness, etc.70 that are inhibitedby karman. The other—pragmatic, as it were—category of results refers to ‘the faculty

Vasudhararakšita/Seó-rgyal 15a.5–15b.4: Kanakavarman/Dad-paÿi œes-rab 95b.5–96a.5:

bya daó bcas par rtogs paÿi phyir |tshad maÿi ÿbras bu ñid du ÿdod ||

bya daó bcas par rtogs paÿi phyir |ÿbras bu ñid du yod tshad ma ||

raó rig la yaó ÿdir ÿbras bu |de yi óo bo las don óes |yul gyi snaó ba ÿdi ñid ÿdi |tshad ma de yis ÿjal bar byed ||

yaó na raó rig ÿdir ÿbras bu |de yi óo bo las don óes |yul gyi snaó ba ñid de ÿdiÿi |tshad ma de yis ÿjal bar bya ||

gaó tshe snaó ba de gšal bya |tshad ma daó deÿi ÿbras bu ni |ÿdsin rnam rig pa de yi phyir |de gsum tha dad du ma byas ||

gaó ltar snaó ba de gšal bya |tshad ma daó deÿi ÿbras bu ni |ÿdsin rnam rig paÿo de yi phyir |de gsum tha dad du ma byas | |

68 tasmât prameyâdhigateÿ sâdhanaô meya-rûpatâ / sâdhane ’nyatra tat-karma-sambandho na prasiddhyati // sâ ca tasyâtma-bhûtÎva tena nârthântaraô phalam /yadâ savišayaô jñânaô jñânâôœe ’rtha-vyavasthiteÿ / tadâ ya âtmânubhâvaÿ saevârtha-viniœcayaÿ //

69 Cf. TS.10.1: moha-kšayâj jñâna-darœanâvaraòântarâya-kšayâc ca kevalam.70 Cf., e.g. TBh.10.7 (p. 231 f.) v.23 ff. (saôsâra-višayâtîtaô muktânâm avyayaô

sukham).

358 PIOTR BALCEROWICZ of appropriation and avoidance’ (âdâna-hâna-dhî) in case all the remaining kinds of(mundane) cognition.71 Consequently, what really the verse is reminiscent of is ratherNBh.1.1.3: yadâ jñânaô tadâ hânôpâdânôpekšâ-buddhayaÿ phalam, with all the threeelements of hâna, upâdana / âdâna and âpekšâ, as a result (phala) of cognition (jñâna).

What is important to remember is that there are numerous similarities, more and lessconspicuous, and not all of them are decisive when taken alone. Some of thesesimilarities indicated on the preceding pages may equally well point to a tradition orauthor prior to Dharmakîrti, viz. to Diónâga or Œaókarasvâmin. Some of suchsimilarities may be due to the general style of writing, of arranging a philosophicaltreatise, of structuring a philosophical discourse, etc. We should remember that boththe Nyâya-bindu and the Nyâyâvatâra were primarily handbooks of logic and theirpurpose was predominantly didactic. Nonetheless, the accumulation of evidence onlyenforces those of them that are quite conclusive and convincing. To sum up, myimpression is that in all dubious cases, when both NP. and NB. seem relevant aspossible sources of Siddhasena’s ideas, Siddhasena probably took recourse toDharmakîrti rather than to Œaókarasvâmin, inasmuch as in all those rare cases whenthere are clear similarities to be found between NA. and NP., they are also traceable inNB. However, not all cases of similarities between NA. and NB. can be shown withregard to NA. and NP. In other words, the development of certain ideas that had takenplace in the period connecting Œaókarasvâmin and Dharmakîrti, was reflected in thecontents of NA. and some ideas still absent from NP., that were later either introducedor modified by Dharmakîrti, found their way into NA. Similarly, certain influences tobe found in NA. point both to Diónâga and to Dharmakîrti. However, Siddhasenaseems to be acquainted with certain new developments or ideas that first developedwith Dharmakîrti (not necessarily only in NB.) and are not found in Diónâga’s works.

Paradoxically as it were, would it not be thinkable to claim that it was Siddhasenawho influenced Dharmakîrti and who was the intermediary stage between Diónâgaand Dharmakîrti? For at least three reasons we should dismiss such a possibility.There is, in the first place, a continuous tradition in epistemic concepts referred toby both Dharmakîrti and Siddhasena that go back to Diónâga. Dharmakîrti himselfrefers to Diónâga so explicitly that would seem highly implausible to believe that hehad availed himself of the Jaina epistemological tradition with respect to the numberof points mentioned above in §§ 1–20, without even a single mention of it: onewould expect Dharmakîrti embarking on at least an accidental discussion of a fewissues he had supposedly taken over from the Jainas. Secondly, Satkari MOOKERJEE,who believed Siddhasena to flourish in the sixth century and to precedeDharmakîrti,72 puzzled over what was in his opinion Dharmakîrti’s lack of reaction

71 Cf. also FRANCO (1997: 65).72 See: MOOKERJEE (1935: 398).

ON THE DATE OF THE NYÂYÂVATÂRA 359 as regards Siddhasena’s interpretation of the intrinsic invariable concomitance(antar-vyâpti) and the superfluous character of the example as an exemplification(bahir-udâhåti) external to the most elementary constituents of the proof formula, aswell as the definition of the logical reason as ‘inexplicability otherwise’.73 Thisbecomes no longer a query when we assume that Siddhasena was post-Dharmakîrtian. Moreover, it is for precisely the same reason that also Pâtrasvâminshould be taken to flourish after Dharmakîrti. Thirdly, the concepts of svârtha-vâkyaand parârtha-vâkya (NA.10) as well as svârtha-pratyakša and parârtha-pratyakša(NA.11) would have with certainty evoked a refutal from the side of Dharmakîrti,had he known about it. Likewise, Dharmakîrti would have certainly commentedupon the idea of non-erroneousness of inference (anumânaô … abhrântaô) provenby its being a cognitive criterion alone (pramâòatvât), found in NA.5. The sameholds true for the idea of ‘inexplicability otherwise’ (anyathânupapannatva). Thus,any supposition that Siddhasena preceded Dharmakîrti can safely be dismissed.

Accordingly, depending on whether we follow the widely accepted dating ofDharmakîrti, viz. c. 600–66074 or the results of latest research by KIMURA (1999) whoassigns the years 550–620 for Dharmakîrti, we would have for the terminus post quemSiddhasena as the author of the Nyâyâvatâra circa 620 or 660, respectively.

There is still another factor to be taken account of, viz. the question of thedefining characteristic of the logical reason (hetu) characterised as ‘the fact of beingotherwise inexplicable’, or ‘inexplicability otherwise’ (anyathânupapannatva,anyathânupapatti) in NA.22: anyathânupapannatvaô hetor lakšaòam îritam. Theauthor clearly refers to an earlier source and the idea did not originate with him inhis NA. Independently, we find the idea reported and criticised by Œântarakšita inTSa.(1).1364 ff. (p. 405 f.) in the context of the validity of inference (anumâna).75

The most famous and relatively often quoted verse is TSa.1369:

anyathânupapannatvaô yatra tatra trayeòa kim /nânyathânupapannatvaô yatra tatra trayeòa kim // 76.

73 See: MOOKERJEE (1935: 4–5): ‘What however strikes us is the intriguing situation

created by Siddhasena’s reference to antarvyâpti and the definition of hetu (probans) asanyathânupapanna in the verse 20 and 22 respectively. It is nothing short of enigma thatthis innovation of the Jaina logicians did not evoke a reply from Dharmakîrti.’

74 See: FRAUWALLNER (1961). Cf. also STEINKELLNER–MUCH (1995: 23).75 The relevant section is edited and translated in KUNST (1939: 11–53). See also

PATHAK (1930–31) 71–83.76 Strangely enough, Œântarakšita in TSa. interchanges the pâdas ab with cd. The verse

is also found in (1) TŒVA. p. 203 [the discussion of anyathânupapatti and the refutation oftri-lakšaòa is found there on pp. 198–217], (2) TBV. Vol. II, p. 569.28–29, (3) PMî.2.1.9

360 PIOTR BALCEROWICZ Significantly enough, Œântarakšita TSa.(1).1364, p. 405.1) mentions Pâtrasvâmin asthe source of the idea: anyathêty-âdinâ pâtrasvâmi-matam âœaókate … , and thetreatise in question is the lost Tri-lakšaòa-kadarthana77 by Pâtrasvâmin, identifiedoccasionally with Pâtrakesarin / Pâtrakesarisvâmin = Vidyânanda. For obviousreasons this Pâtrasvâmin cannot be Vidyânanda (c. 850), the author of the Œloka-vârttika on Umâsvâti’s Tattvârtha-sûtra—as SUKHLAL–DOSHI (1928)78 andCHATTERJEE (1978: 331) would have it—but some else who preceded Œântarakšita(c. 725–788)79, the teacher of Kamalaœîla.80 Since the author of NA. alludes to hispredecessors and Œântarakšita mentions only Pâtrasvâmin, it must have been thelatter who was responsible for the idea of anyathânupapannatva.81 Had it been

§ 33 (p. 45.17–18). Hemacandra’s criticism against the Buddhist idea of trairûpya inPMî.2.1.9 § 33 (p. 45.1–16) closely follows the exposition of Pâtrasvâmin’s aphorismsquoted in TSa. attesting to the authenticity of the quotation. Hemacandra, instead of theterms anyathânupapatti, uses the expression avinâ-bhâva, cf. PMî.2.1.9 (p. 43.34–35):svârthaô sva-niœcita-sâdhyâvinâ-bhâvÎka-lakšaòât sâdhanât sâdhya-jñânam. Theformulation sâdhyâvinâ-bhâvÎka-lakšaòât resembles both Pâtrasvâmin’s Tri-lakšaòa-kadarthana (tenÎka-lakšaòo hetuÿ prâdhânyâd gamako ’stu naÿ / = TSa.1379) as well asNA.5ac: sâdhyâvinâ-bhuno liógât sâdhya-niœcâyakaô småtam / anumânaô.

77 A reference to the work is found in DHAKY (1995: 43), who refers to Jugal KishorMukhtar: ‘Saômatisûtra aur Siddhasena’ (Hindi), Jaina Sâhitya aur Itihâsa par ViœadaPrakâœa, Calcutta 1956: 538–543 [the work was not available to me].

78 Cf. UPADHYE (1971: *14–15), PATHAK (1930: passim) and PATHAK (1930–31:passim), who refers to him as Pâtrakesari Vidyânanda or as Pâtrakesarisvâmi.

79 Cf. STEINKELLNER–MUCH (1995: 56).80 Cf. BHATTACHARYYA (1926: ixvi-ixvii): ‘In that case Pâtrasvâmin must be an

earlier author than both Œântarakšita and Vidyânanda, and he must have firstpropounded the theory that valid reason is that the existence of which cannot bemaintained unless it is invariably concomitant with the major term … cir. 700 A.D.’.

81 Cf. MUKTHAR (1948) [according to UPADHYE (1971:*30)]: ‘7) The Nyâyâvatâra iscomposed centuries later than the Sanmati-sûtra, because it shows the influence ofPâtrasvâmi (later than Samantabhadra) as well as Dharmakîrti and Dharmottara’; andV.P. Johrapurkar (‘Introduction’ to VTP., pp. 41 ff.) quoted in n. 9 above. See alsoDHAKY (1995: 42–3): ‘…the first foot of the kârikâ 22 concerning the hetu-lakšaòa(character of probans) reflects sense-agreement, even partial verbal concordance withthe verses from Pâtrasvâmi’s (Pâtrakesari’s) Trilakšaòakadarthana cited by theBuddhist scholiast Œântarakšita in his Tattvasaógraha (c. 2nd quarter of the 8th century A.D.)32.’ SHASTRI (1990: 31) is a bit more reserved and does not take for granted thatPâtrasvâmin was anterior to Siddhasena: ‘Nyâyâvatâra also defines hetu in the samemanner.’

ON THE DATE OF THE NYÂYÂVATÂRA 361 Siddhasena who introduced the idea, Œântarakšita would not, in all probability, havemissed the opportunity to mention this. Certainly Pâtrasvâmin is post-Diónâgan, forhis Tri-lakšaòa-kadarthana was conceived to refute the latter, but his dating is quiteuncertain. It is surprising, nonetheless, that Œântarakšita seems nowhere to allude tothe NA. or its author. Consequently, the widespread opinion assuming that NA. isthe first Jaina treatise on epistemology per se loses its weight, inasmuch we cansafely assume that Siddhasena had his predecessor in the person of Pâtrasvâmin.

As regards the terminus ante quem, in view of the fact that the verse no. 4 of theNyâyâvatâra is incorporated into ŠDSa. as verse no. 56, it should be assigned to thedate of Haribhadrasûri. A supposition that it is ŠDSa.56 that was the source whichNA.4 was borrowed from and that NA.4 is merely an interpolation seemsinadmissible to me. NA.4 fits ideally the argumentative structure of the text:(1) NA.1 states the definition of pramâòa, as well as the types and a generalcriterion of such a division, (2) NA.2 is polemical concerning the meaninglessnessof formulating a definition for a well-known term / idea of pramâòa, (3) NA.3 is arejoinder to the objection, (4) NA.4 defines the two main divisions of pramâòaalong with a specific reason for such a division,82 (5) NA.5 opens a section dealingwith erroneousness of pramâòas, etc. The use of vocabulary in NA.4 is not unusualfor NA. Coincidentally, two adjacent verses, viz. NA.4 and NA.5, are stylisticallyand structurally quite akin: pâda a–b: €reason� (aparokšatayâ, sâdhyâvinâ-bhunoliógât) + €object + verbal derivative in the meaning of a present participle�(arthasya grâhakaô, sâdhya-niœcâyakaô) + … + pâda c: €subject� (pratyakšam,anumânaô) … Admittedly, the similarity is not a crucial argument in favour of thesame authorship of the two verses, however, there is nothing that could speakagainst a common authorship. The argument gains on strength in view of the factthat Haribhadrasûri quotes the verse no. 2 of NA. in his Ašþaka83 and refers to itsauthor as Mahâmati84.

82 Both NA.1 and 4 go against the prevalent Jaina tradition to subsume cases of

sensory cognition, inference and verbal testimony under parokša, whereas pratyakšawas taken to denote extra-sensory and extra-mental acts of cognition (viz. avadhi,manaÿ-paryâya and kevala). For Siddhasena the criterion of directness (akša) was notthe cognising subject, or the soul (akša=jîva=âtman), but—like in the general pramâòatradition—the senses (akša=indriya).

83 The work is not available to me. I am forced to rely here on Pt. DalsukhbhaiMALVANIA (1979: 287–288). Cf. also UPADHYE (1971: xxiv) and DHAKY (1995: 44).

84 UPADHYE (1971: xxiv) is right to point out ‘that Haribhadra, in his Ašþaka, quotesthe Nyâyâvatâra 2, by referring to its author as Mahâmati. Elsewhere, howeverHaribhadra speaks plainly about the author of the Sanmati as Divâkara andŒrutakevalin.’


In view of the above, DHAKY’s (1995: 44) claim85—to handle the discomfort thatHaribhadra himself ascribes one of the verses to a Mahâmati—that both NA.2 and 4were taken from lost dvâtriôœikâs of Siddhasena Divâkara is highly debatable tome. Not only NA.4 seems to represent an original kârikâ of NA., but the same holdsgood in the case of kârikâ 2. We could not make head or tail of NA.2 (the objection)if we did not have NA.1. Moreover, NA.3 (the rejoinder) would be pointless withoutNA.2. All the lost dvâtriôœikâs of Siddhasena Divâkara as a reference source inargumentation resemble rather a kind of œaœa-višâòa. Naturally, as long as we donot have all dvâtriôœikâs written by Siddhasena Divâkara, we cannot, theoreticallyspeaking, preclude the possibility that the verses indeed were taken from some lostdvâtriôœikâ. But such an argumentation is not very constructive. I see no reason todoubt the authenticity of kârikâs 2 and 4 as long as they form a consistent logicalpart of the Nyâyâvatâra as a whole and bear stylistic similarities to adjacentkârikâs,86 especially when the counter-arguments begin with ‘perhaps’ and are ofmerely could-or-may-have-been nature.

Further, the identification of Siddharšigaòi as the author of both NA. and NAV.,postulated by DHAKY, has very weak foundations. In the first place, there is ampleevidence that Siddharšigaòi (the author of NAV.) is not Siddhasena (the author of

85 ‘A formidable objection, however, to the above-postulated identification (Siddharši

was the author of both NA. and NAV.—P.B.) as well as the period determination(ninth/tenth century for NA.—P.B.) can be raised on the grounds of the ascription of averse, which appears as the kârikâ 2 of the Nyâyâvatâra, to Mahâmati (=SiddhasenaDivâkara) by Yâkinisûnu Haribhadra sûri (active c. A. D. 745–785) in his Ašþaka38. Andthe kârikâ 4 figures as a part of the Šaðdarœana-samuccaya of the same Haribhadrasûri39. Since Haribhadra sûri ascribed the particular verse (kârikâ 2) to SiddhasenaDivâkara, it must be so. However, this kârikâ could be originally from somedvâtriôœikâ, one of the lost 11 of Siddhasena Divâkara, perhaps the Pramâòa-dvâtriôœikâ, from which Gandhahasti Siddhasena quotes in his sa-bhâšya-Tattvârthâdhigama-sûtra-våtti40. The kârikâ 4 in the Šað-darœana-samuccaya maylikewise have been taken from one of the unavailable dvâtriôœikâ of SiddhasenaDivâkara. Alternatively, if that verse is Haribhadra’s own, Siddharši must haveborrowed it from the Šaðdarœana. In any case, Haribhadra and Siddharši could havecommon sources from which they apparently may have drawn.’

86 I have dealt briefly with the style of NA.4 above. The same is even more true forNA.2 that is closely followed by NA.3 (the most conspicuous similarities underlined):

/2/ prasiddhâni pramâòâni vyavahâraœ ca tat-kåtaÿ /pramâòa-lakšaòasyôktau jñâyate na prayojana m //

/3/ prasiddhânâô pramâòânâô lakšaòôktau prayojana m /tad-vyâmoha-nivåttiÿ syâd vyâmûðha-manasâm iha //

ON THE DATE OF THE NYÂYÂVATÂRA 363 NA.) and that the two texts were written by different authors, inasmuch asSiddharšigaòi refers to the author of NA. explicitly, although not by name, but bythe term âcârya or sûtra-kåt, to cite a few cases only:87 [1] Introductory lines ofNAV.3: /3/ adhunâcâryo gåhîtas tâvakîno ’bhiprâyo ’smâbhir iti paraôpratyâyayaôs tan-matam anudrâvya tad evânumanyamânas tathâpi lakšaòôkteÿsâphalyam âvedayann âha: …; [2] the final sentence of NAV.3: tad evaô pramâòa-lakšaòaô sâmânyena pratipâdya tad-gataô kucodyaô paryahâry âcâryeòa;[3] NAV.13: yad vâtyantâbhyâsena parikarmita-matitvât tâvatÎva prastuta-prameyam avabudhyate, tadâ hetu-pratipâdanam eva kriyate, œešâbhidhânasyaœrotå-saôskârâkâritayâ nairarthakyâd ity-âdau hetu-pratipâdanaô sûtra-kåtâparârtham anumânam uktam; [4] NAV.29: ata evâcâryasya na tal-lakšaòâdi-svarûpa-kathane ’pi mahânâdaraÿ. Additionally, in some cases (e.g. on NA.8),Siddharšigaòi does offer at least two different interpretations of a kârikâ, or divergesfrom the contents of NA., e.g. in the case of NA.4d88.

Thus, we can safely take the date-brackets for the Nyâyâvatâra to be firmly fixedafter 620 C.E. (Dharmakîrti) and Pâtrasvâmin and before c. 800 C.E.(Haribhadrasûri).89 As for the date of Haribhadrasûri, JACOBI (1926: Introduction)assigns Haribhadra to c. 750, whereas UPADHYE (1971: xxv) to c. 750–800 C.E. andDHAKY (1995: 44) to c. 745–785. However, the terminus post quem for Haribhadrais Arcaþa, since the latter is quoted in Haribhadrasûri’s NPV.9.15–1990. Arcaþa, theteacher of Dharmottara, can be assigned to c. 730–790 or 720–780.91

87 DHAKY (1995: 43) is so far right that indeed Siddharši nowhere mentions the

author of Nyâyâvatâra by name: ‘As noted in the beginning, Siddharši does not ascribethe Nyâyâvatâra to Siddhasena Divâkara or to a different Siddhasena or for that matterto any other author.’ However, he clearly misses the point when he further claims: ‘Nordoes he mention it as a composition of a pûrvâcârya, våddhâcârya, or somecirantanâcârya. Also, in his verse by verse exposition, he nowhere uses qualificatoryphrases such as œâstrakâra, sûtrakâra, kârikâkâra, âcârya, etc. which may have denoteda second, an earlier revered personage, as the kârikâs’ author.’

88 Siddhasena emphasises the way of cognising, or ‘[the manner of] grasping [anobject]’, whereas Siddharši takes the expression to refer to the existence and the natureof the cognoscible. See p. 341 above.

89 Thus, I cannot but side with the opinion already expressed in VAIDYA (1928: xx):‘The terminus a quo would be the date of Dharmakîrti and the terminus ad quem that ofHaribhadra.’

90 Acc. to STEINKELLNER–MUCH (1995: 119) the original title found to the colophonis Nyâya-praveœa-þîkâ œišya-hitâ.

91 See: STEINKELLNER–MUCH (1995: 64) and FRAUWALLNER (1961: 148).


BIBLIOGRAPHY

AJP. = Haribhadra Sûri: Anekânta-jaya-patâkâ. Ed. with his owncommentary and Municandra Sûri’s supercommentary. Vol. 1 & 2,Ed. H. R. Kâpadîâ, Gaekwad’s Oriental Series 88, 105, OrientalInstitute, Baroda 1940, 1947.

BALCEROWICZ 1995 = Balcerowicz, Piotr: ‘Œaókarasvâmin: Nyâya-praveœa—“Wprowadzenie w logikê” (“Introduction to Logic”).’ Part One:I. Polish Translation, II. Sanskrit Text, III. Notes § I, Abbreviationsand Bibliography. Studia Indologiczne (Warsaw) 2 (1995) 39–87.

BALCEROWICZ 1999 = Balcerowicz, Piotr: ‘Taxonomic approach to dåšþântâbhâsa inNyâya-bindu and in Siddharšigaòi’s Nyâyâvatâra-vivåti—Dharmakîrti’s typology and the Jaina criticism thereof.’ In:Dharmakîrti’s Thought and Its Impact on Indian and TibetanPhilosophy (Proceedings of the Third International DharmakîrtiConference—Hiroshima, November 4–6, 1997). Ed. Shoryu Katsura,Verlag der Österreichischen Akademie der Wissenschaften: Beiträgezur Kultur- und Geistesgeschichte Asiens Nr. 32, Wien 1999: 1–16.

BALCEROWICZ 2001 = Balcerowicz, Piotr: ‘Two Siddhasenas and the Authorship of theNyâyâvatâra and the Saômati-tarka-prakaraòa’. Journal of IndianPhilosophy 29/3 (2001) 351–578.

BALCEROWICZ(forthcoming)

= Balcerowicz, Piotr: ‘On the Relationship of the Nyâyâvatâra andthe Saômati-tarka-prakaraòa.’ Presented at the XIth WorldSanskrit Conference, Turin, April 3rd–8th, 2000; forthcoming.

BHATTACHARYYA 1926 = Bhattacharyya, B.: ‘Foreword’ to TSa.(1).i-cxvii.BSŒBh. = Œaókara: Brahma-sûtra-œâókara-bhâšya. Ed. with the

commentaries: Bhâšya-ratna-prabha of Govindânanda, Bhâmatî ofVâcaspatimiœra, Nyâya-niròaya of Ânandagiri; ed. by J.L. Shastri,Motilal Banarsidass, Delhi 1980. [Reprinted: Delhi 1988].

CHATTERJEE 1978 = Chatterjee, Asim Kumar: A Comprehensive History of Jainism (upto 1000 A.D.), Firma KLM Private Ltd., Calcutta 1978.

DAVE 1962 = Dave, P. N.: Siddhasena Divâkara: A Study. A Thesis submitted tothe University of Bombay for the Ph.D. degree, September 1962.[After UPADHYE (1971:*51)].

DHAKY 1981–82 = Dhaky, M. A.: ‘Some less known verses of Siddhasena Divâkara.’Sambodhi 10 (1981–82) 169–173.

DHAKY 1990 = Dhaky, M. A.: ‘The Date and Authorship of Nyâyâvatâra.’ Atypescript of a paper—presented to me by the author in 1990—prepared for publication in B. K. Thapar Felicitation Volume [theedition was not available to me].

DHAKY 1995 = Dhaky, M. A.: ‘The Date and Authorship of Nyâyâvatâra.’Nirgrantha (Ahmedabad, eds. M. A. Dhaky, Jitendra Shah) 1(1995) 39–49.

ON THE DATE OF THE NYÂYÂVATÂRA 365 DhPr. = Paòðita Durveka Miœra’s Dharmottara-pradîpa [being a

subcommentary on Dharmottara’s Nyâya-bindu-þîkâ, a commentaryon Dharmakîrti’s Nyâya-bindu], Ed. by Pt. Dalsukhbhai Malvania,Kashi Prasad Jayaswal, Research Institute, Patna 1971.

FRANCO 1987 = Franco, Eli: Perception, knowledge and disbelief: a study ofJayarâœi’s scepticism. Alt- und Neu-Indische Studien 35, FranzSteiner Verlag, Stuttgart 1987. [Reprinted: Motilal Banarsidass,Delhi 1994].

FRANCO 1997 = Franco, Eli: Dharmakîrti on Compassion and Rebirth. WienerStudien zur Tibetologie und Buddhismuskunde 38, Wien 1997.

FRAUWALLNER 1933 = Frauwallner, Erich: ‘Zu den Fragmenten buddhistischer Logiker imNyâyavârttikam.’ Wiener Zeitschrift für die Kunde Morgenlandes40 (1933) 300–304. [Reprinted: FRAUWALLNER (1982: 479–483)].

FRAUWALLNER1957/a

= Frauwallner, Erich: ‘Vasubandhu’s Vâdaviddhi.’ WienerZeitschrift für die Kunde Morgenlandes (1957) 104–146.[Reprinted: FRAUWALLNER (1982: 716–758)].

FRAUWALLNER1957/b

= Frauwallner, Erich: ‘Review of R. Sâókåtyâyana’s Edition of thePramâòavârtikabhâšya.’ Journal of the American Oriental Society(1957) 58–60. [Reprinted: FRAUWALLNER (1982: 883–885)].

FRAUWALLNER 1961 = Frauwallner, Erich: ‘Landmarks in the History of Indian Logic.’Wiener Zeitschrift für die Kunde Süd- und Ostasiens 5 (1961) 125–148.

FRAUWALLNER 1982 = Frauwallner, Erich: Kleine Schriften. Hrsg. von GerhardOberhammer und Ernst Steinkellner, Glassenapp-Stiftung 22,Franz Steiner Verlag, Wiesbaden 1982.

GRANOFF 1989–1990 = Granoff, Phyllis: ‘The Bibliographies of Siddhasena—A Study inthe Texture of Allusion and the Weaving of a Group Image.’ PartI: Journal of Indian Philosophy 17 (1989) 329–384. Part II:Journal of Indian Philosophy 18 (1990) 261–304.

HATTORI 1968 = Hattori, Masaaki: Dignâga, On Perception, being the Pratyakša-pariccheda of Dignâga’s Pramâòa-samuccaya, edition of Tibetantranslations and the Sanskrit text as well as the English translationof the Chapter I. Harvard University Press 1968.

HB. = Steinkellner, Ernst: Dharmakîrti’s Hetubinduÿ. Teil I: TibetischerText und rekonstruierter Sanskrit-Text; Teil II: Übersetzung undAnmerkungen. Veröffentlichungen der Kommission für Sprachenund Kulturen Süd- und Ostasiens 4, 5; Wien 1967.

JACOBI 1926 = Jacobi, Hermann: ‘Introduction and Contents’ to Haribhadra:Samarâïcca-kahâ—A Jaina Prakåta Work. Volume I: Text andIntroduction, Bibliotheca Indica, Calcutta 1926: i-cxxx.

KIMURA 1997 = Kimura, Toshihiko: ‘New Chronology of Dharmakîrti’, in: In:Dharmakîrti’s Thought and Its Impact on Indian and TibetanPhilosophy (Proceedings of the Third International DharmakîrtiConference—Hiroshima, November 4–6, 1997). Ed. by Shoryu


Katsura, Verlag der Österreichischen Akademie derWissenschaften: Beiträge zur Kultur- und Geistesgeschichte AsiensNr. 32, Wien 1999: 209–214.

KRAUSE 1948 = Krause, Charlotte: ‘Siddhasena and Vikramâditya.’ VikramaVolume, Scindia Oriental Institute, Ujjain 1948: 213–280.

KUNST 1939 = Kunst, Arnold: Probleme der buddhistischen Logik in derDarstellung des Tattvasaógraha (Zagadnienia logiki buddyjskiejwed³ug Tattvasangrahy Siantarakszity). Polska AkademiaUmiejêtnoœci, Mémoires de la Commission Orientaliste 33,Kraków 1939.

MALVANIA 1979 = Pt. Dalsukhbhai Malvania: ‘Pariœišþa 1, Nyâyâvatara kî Tulanâ’.1979.

MATILAL 1985 = Matilal, Bimal Krishna: Logic, Language and Reality: IndianPhilosophy and Contemporary Issues. Motilal Banarsidass, Delhi1985.

MAV. = Asaóga (Maitreyanâtha?): Madhyânta-vibhâga[-sûtra]; ed. byNathmal Tatia and Anantalal Thakur with Vasubandhu’s Bhâšya,Tibetan Sanskrit Works Series 10, K.P. Jayaswal ResearchInstitute, Patna 1967.

MAVBh. = Vasubandhu: Madhyânta-vibhâga-bhâšya. See: MAV.Met. = Aristotelis Metaphysica. Ed. W. Jaeger, Scriptorum Classicorum,

Bibliotheca Oxoniensis, Oxford University Press, London 1957.MOOKERJEE 1935 = Mookerjee, Satkari: The Buddhist Philosophy of Universal Flux –

An Exposition of the Philosophy of Critical Realism as Expoundedby the School of Dignâga. Motilal Banarsidass, Delhi 1993. [Firstedition: University of Calcutta, 1935].

MOOKERJEE 1971 = Mookerjee, Satkari: ‘A critical and comparative study of Jain Logicand Epistemology on the basis of the Nyâyâvatâra of SiddhasenaDivâkara.’ Vaishali Institute Research Bulletin 1, Vaishali 1971.

MSA. = Asaóga (Maitreyanâtha?): Mahâyâna-sûtrâlaókâra. (1) Ed. SylvainLévi, Paris 1907. (2) Ed. and transl. Surekha Vijay Limaye, BIBS 94,Sri Satguru Publications / Indian Books Centre, Delhi 1992.

MŒV. = Kumârila Bhaþþa: Mîmâôsâ-œloka-vârttika. Œlokavârttika of ŒrîKumârila Bhaþþa with the Commentary Nyâya-ratnâkara of ŒrîPârthasârathi Miœra. Ed. Dvârikâdâsa Œâstrî, Ratnabharati Series3, Târâ Publications, Varanasi 1978.

MUKTHAR 1948 = Mukthar, Jugalkishore: ‘Sanmati Siddhasenâóka.’ Anekânta,IX.11–2, December 1948; reprinted in the Introduction of thePurâtana-Jaina-Vâkya-sûcî, Delhi 1950. Included in his JainaSâhitya aura Itihâsapara Viœada-prakâœa, Part I, Delhi 1956; itsEnglish Translation (by A. N. Upadhye): Sanmati-sûtra andSiddhasena, Virasewamandira Trust, Delhi 1965. [After UPADHYE(1971:*25)].

NA. = Siddhasena Divâkara: Nyâyâvatâra. (1) Ed. with Siddharšigaòi’s


Vivåti and Devabhadra’s Þippana, by Bhagavandas Harakhchand;Ahmedabad–Patan 1917. (2) Ed. with Siddharšigaòi’s Vivåti andDevabhadra’s Þippana, by P. L. Vaidya, Shri Jain ShwetamberConference, Bombay 1928. [Reprinted in Vaishali InstituteResearch Bulletin 1, Vaishali 1971: 1–95].

NASV. = Nyâyâvatâra-sûtra-vârttika of Œrî Œânti Sûri critically edited inSanskrit with notes, indices etc. in Hindi [with the Våtti] byDalsukh Malvania, published by Singhi Jain Œastra Œikshapitha,Bharatiya Vidya Bhavan, Bombay 1949.

NAÞ. = Devabhadra’s Nyâyâvatâra-þippana. See: NA.NAV. = Siddharšigaòi: Nyâyâvatâra-vivåti. See: NA.NAVV. = Œrî Œânti Sûri: Nyâyâvatâra-vârttika-våtti. See: NASV.NB. = Dharmakîrti: Nyâya-bindu. See: DhPr.NBh. = Vâtsyâyana Pakšilasvâmin: Nyâya-bhâšya. See: NS.NBÞ. = Dharmottara: Nyâya-bindu-þîkâ. See: DhPr.NM. = Nyâya-mukha of Dignâga. Ed. Giuseppe Tucci, Heidelberg 1930.NP. = Œaókarasvâmin: Nyâya-praveœa. (1) [Part One:] Nyâya-praveœa of

Diónâga with Commentaries of Haribhadra Suri & Parsavadeva’ed. by A. B. Dhruva, Oriental Institute, Baroda 1930. (2) See:BALCEROWICZ (1995: 72–77).

NPV. = Haribhadrasûri: Nyâya-praveœa-våtti. See: NP.(1).NS. = Akšapâda Gautama: Nyâya-sûtra. Mahâmuni Vâtsyâyan’s Nyâya-

bhâšya on Gautama-Nyâya-sutras With PrasannapadâCommentary of Pt. Sudarœanâcârya Shastrî. Ed. DvârikâdâsaŒâstrî, SS 10, Sudhî Prakâœanam, Vârâòasî 1986.

NV. = Uddyotakara: Nyâya-vârttika. Nyâya-darœana (= Nyâya-sûtra)with Vâtsyâyana’s Bhâšya, Uddyotakara’s Vârttika, VâcaspatiMiœra’s Tâtparyaþîkâ and Viœvanâtha’s Våtti. ed. by TaranathaNyaya-Tarkatirtha and Amarendramohan Tarkatirtha, Kalikâtâ-saôskåta-grantha-mâlâ 18, Calcutta 1936.

PATHAK 1930 = K.B. Pathak: ‘Œântarakšita’s reference to Kumârila’s Attacks onSamantabhadra and Akalaókadeva.’ Annals of the BhandarkarOriental Research Institute 11 (1930) 155–164.

PATHAK 1930–31 = K.B. Pathak: ‘Dharmakîrti’s Trilakšaòahetu attacked byPâtrakesari and defended by Œântarakšita.’ Annals of theBhandarkar Oriental Research Institute 12 (1930–31) 71–83.

PBh. = Praœastapâda: Praœastapâda-bhâšya. Bronkhorst, Johannes andRamseier, Yves: Word index to the Praœastapâdabhâšya: acomplete word index to the printed editions of thePraœastapâdabhâšya; Motilal Banarsidass, Delhi 1994.

PMî. = Hemacandra: Pramâòa-mîmâôsâ. Ed. and trans. by SatkariMookerjee and Nathmal Tatia, Târâ Publications, Varanasi 1970.

PS. = Diónâga: Pramâòa-samuccaya. (1) Chapter I, see: HATTORI


(1968). (2) The Tibetan Tripitaka—Peking Edition, ed. Daisetz T.Suzuki, Tibetan Tripitaka Research Institute, Tokyo—Kyoto1957—Bstan-ÿgyur, Mdo-ÿgrel, Gtan-Tshogs Rig-pa I, Vol. 130,No.5700.

PV. = The Pramâòa-vârttikam of Âcârya Dharmakîrti with Sub-commentaries: Svôpajña-våtti of the Author and Pramâòa-vârttika-våtti of Manorathanandin. Ed. by Prof. Dr. Ram Chandra Pandeya,Motilal Banarsidass, Delhi 1989.

PVin.I(1) = Dharmakîrti’s Pramâòâviniœcayaÿ, 1. Kapitel: pratyakšaô. Hrsg.und übers. von T. Vetter. Veröffentlichungen der Kommission fürSprachen und Kulturen Süd- und Ostasiens Heft 3, ÖsterreichischeAkademie der Wissenschaften, Wien 1966.

PVin.II(1) = Dharmakîrti’s Pramâòâviniœcayaÿ, 2. Kapitel: Svârthânumânaô.Tib. Text und Sanskrittexte von Ernst Steinkellner.Veröffentlichungen der Kommission für Sprachen und KulturenSüd- und Ostasiens 12, Österreichische Akademie derWissenschaften, Wien 1973.

PVin.II(2) = Dharmakîrti’s Pramâòâviniœcayaÿ, 2. Kapitel: Svârthânumânaô.übers. von Ernst Steinkellner. Veröffentlichungen der Kommissionfür Sprachen und Kulturen Süd- und Ostasiens 15, ÖsterreichischeAkademie der Wissenschaften, Wien 1979.

PVSV. = Dharmakîrti: Pramâòa-vârttika-svôpajña-våtti. See: PV.PVV. = Manorathanandin: Pramâòa-vârttika-våtti. See: PV.QVARNSTRÖM 1999 = Qvarnström, Olle: ‘Haribhadra and the Beginnings of Doxography

in India.’ In: Approaches to Jaina Studies: Philosophy, Logic,Rituals and Symbols [Proceedings of the International Conferenceon Approaches to Jaina Studies: Philosophy, Logic, Rituals andSymbols, 31.03–2.04.1995 Toronto]. Ed. by N.K. Wagle and OlleQvarnström, South Asian Studies Papers 11, University ofToronto: Centre for South Asian Studies, Toronto 1999: 169–210.

RANDLE 1926 = Randle, H.N.: Fragments from Dignâga. Prize Publication Fund,Royal Asiatic Society, London 1926. [Reprinted: MotilalBanarsidass, Delhi 1981].

SacAcBh. = Asaóga: Sacittikâ Acittikâ ca Bhûmiÿ [A Section of Yogâcâra-bhûmi]. (1) Wayman, Alex (Ed.): ‘The Saccittikâ And AcittikâBhûmi—Text and Translation.’ In: Buddhist Insight—Essays byAlex Wayman. Ed. with an Introduction by George Elder, BuddhistTradition Series VII, Delhi 1984. [Reprinted Motilal Banarsidass,Delhi 1990; 1st ed.: Journal of Indian and Buddhist Studies(Tokyo) 7/1 (1960) 375–379]. (2) See: SCHMITHAUSEN (1987:220–2, ‘Appendix I’).

SCHMITHAUSEN 1987 = Schmithausen, Lambert: Âlayavijñâna—On the Origin and theEarly Development of a Central Concept of Yogâcâra Philosophy.Part I: Text, Part II: Notes, Bibliography and Indices. StudiaPhilologica Buddhica, Tokyo IVa/IVb, The International Institute


for Buddhist Studies, Tokyo 1987.ŠDSa. = Haribhadrasûri: Šað-darœana-samuccaya. Ed. with the

Commentaries of Tarka-rahasya-dipikâ of Guòaratnasûri andLaghuvåtti of Somatilaka Sûri and an Avacûròi; by MahendraKumar Jain, with the Introduction of Pt. Dalsukh Malvania, JMJG-SG 36, Benares 1981.

SHASTRI 1990 = Shastri, Indra Chandra: Jaina Epistemology. P. V. Research SeriesNo. 50, P. V. Research Institute, Varanasi 1990.

STEINKELLNER–KRASSER 1989

= Steinkellner, Ernst; Krasser, Helmut: Dharmottaras Exkurs zurDefinition gültiger Erkenntnis im Pramâòa-viniœcaya. Beiträge zurKultur- und Geistesgeschichte Asiens No.2, ÖsterreichischeAkademie der Wissenschaften, Wien 1989.

STEINKELLNER–MUCH 1995

= Steinkellner, Ernst; Much, Michael Torsten: Texte dererkenntnistheoretischen Schule des Buddhismus—SystematischeÜbersicht über die buddhistische Sanskrit-Literatur II. Abh. derAkad. der Wissenschaften in Göttingen, Philologisch-HistorischeKlasse, Dritte Folge 214, Vandenhoeck & Ruprecht, Göttingen1995.

SUKHLAL 1945/a = Pt. Sukhlalji Sanghavi: ‘Pratibhâmûrti Siddhasena Divâkara.’Bhâratîya Vidyâ, III.i, Bombay 1945: 9–20 (in Hindi) [See:UPADHYE (1971:*18)].

SUKHLAL 1945/b = Pt. Sukhlalji Sanghavi: ‘Œrî Siddhasena Divâkarano SamayanoPraœna.’ Bhâratîya Vidyâ, III, Bombay 1945: 152 f. (in Gujarati)[See: UPADHYE (1971:*21)].

SUKHLAL–DOSHI1928

= Pt. Sukhlalji Sanghavi; Bechardas Doshi: ‘Sanmati-tarka and itsImportance.’ Jaina Silver Jubilee Number, Bhavnagar 1928 (Saô.1985): 109–121 (in Gujarati) [After UPADHYE (1971:*12)].

TABER 1994 = Taber, A. John: ‘Kumârila’s Refutation of the DreamingArgument: the Nirâlambanavâda-adhikaraòa.’ In: Studies inMîmâôsâ—Dr. Mandan Mishra Felicitation Volume. Ed. R. C.Dwivedi, Motilal Banarsidass, Delhi 1994: 27–52.

TBh. = Umâsvâti: Tattvârthâdhigama-bhâšya. See TS.TBV. = Abhayadevasûri’s Tattva-bodha-vidhâyinî on Siddhasena

Divâkara’s Saômati-tarka-prakaraòa. Ed. by Sukhlâl Saóghaviand Becardâs Doœi; Vol. I & II, Rinsen Buddhist Text Series VI1,2; Kyoto 1984. [Reprinted from the original edition published in5 Vols., Gujarât-purâ-tattva-mandir-granthâvalî 10, 16, 18, 19,21, Gujarât-purâ-tattva-mandir, Amdâvâd, 1924–1931].

TILLEMANS 1997 = Tillemans, Tom J.F.: ‘Dharmakîrti on prasiddha and yogyatâ.’ In:Aspects of Buddhism: Proceedings of the International Seminar onBuddhist Studies, Liw, 25 June 1994, eds.: Agata BarejaStarzyñska, Marek Mejor = Studia Indologiczne (Warszawa) 4(1997) 177–194.

Triôœ. = Vasubandhu: Triôœikâ-vijñapti-kârikâ. See: VMS.

370 PIOTR BALCEROWICZ TS. = Umâsvâmin: Tattvârtha-sûtra. Ed. together with

Tattvârthâdhigama-bhâšya by M.K. Mody, BI No.1044, Calcutta1903–5.

TSa. = Œântarakšita’s Tattva-saógraha together with Kamalaœîla’sPañjikâ. Ed. by Embar Krishnamacharya with a Foreword by B.Bhattacharyya, 2 Vols., Gaekwad’s Oriental Series 30–31, Baroda1926. [Reprinted: 1984].

TSaP. = Kamalaœîla: Tattva-saógraha-pañjikâ. See: TSa.

TŒVA. = Vidyânanda Pâtrakesarisvâmin: Tattvârtha-œloka-vârtikâlaôkara.Ed. by Manoharalâl, Ramacandra Nâtha Raógaji, Bombay 1918.

UPADHYE 1971 = Upadhye, A. N.: Siddhasena Divâkara’s Nyâyâvatâra (edited bythe late S.C. Vidyabhusan with English Translation, Notes etc.)and with the Vivåti of Siddharši as well as The Text of 21Dvâtriôœikâs and the Sammaï-suttam; Vinayavijaya’s Nayakaròikâedited by … with an Introduction, Bibliographic Review, Indicesetc.; Jaina Sâhitya Vikâsa Maòðala, Bombay 1971.

VAIDYA 1928 = Vaidya, P. L.: ‘Introduction’ to Nyâyâvatâra. See: NA.(2): vii-xliii.

Viôœ. = Vasubandhu: Viôœatikâ. See: VMS.

VMS. = Vasubandhu: Vijñapti-mâtratâ-siddhi (Viôœatikâ and Triôsikâ).Deux traités de Vasubandhu, Viôœatikâ et Triôœikâ, 1 Partie—Texte, Sthiramati’s Bhâšya. Ed. by Sylvain Lévi, Bibliothèque del’École des Hautes Études, Paris 1925.

VS. = The Vaiœešika-sûtra of Kaòâda with the Commentary ofCandrânanda. Ed. Muni Jambûvijayaji, GOS 136, Baroda 1961.

VTP. = Bhâvasena: Viœva-tattva-prakâœa. Ed. by Vidyadhara PasusaJohrapurkar, Jîvarâja Jaina Granthamâlâ, Sholapur 1964.

WAYMAN 1996 = Wayman, Alex: ‘The Nyâyâvatâra and Buddhist Logical Works byDignâga and Ratnâkaraœânti.’ Nirgrantha 2 (1996) 1–8, eds. M. A.Dhaky, Jitendra Shah, Ahmedabad.

WILLIAMS 1963 = Williams, Richard: Jaina Yoga. London Oriental Series 14, NewYork–Toronto 1963.

on the date of the nyâyâvatâra - piotr balcerowicz · 2016-11-07 · on the date of the...

Documents